diff options
| author | Paul Kirth <paulkirth@google.com> | 2024-07-02 08:24:03 -0700 |
|---|---|---|
| committer | Paul Kirth <paulkirth@google.com> | 2024-07-02 08:24:03 -0700 |
| commit | 803a88c3bcf182072533186d3bb07716fe8ea8cd (patch) | |
| tree | 3859a561fed884f6981ca1a071f6123b984a3d39 | |
| parent | 9bfd2224f4d783ffe74ded69c4f8d15a70df5608 (diff) | |
| parent | ac7b1d0cf1ff30183bb9487fbfef3c4e48bbbf53 (diff) | |
| download | llvm-users/ilovepi/spr/reapply-llvmriscv-enable-trailing-fences-for-seq-cst-stores-by-default-87376.zip llvm-users/ilovepi/spr/reapply-llvmriscv-enable-trailing-fences-for-seq-cst-stores-by-default-87376.tar.gz llvm-users/ilovepi/spr/reapply-llvmriscv-enable-trailing-fences-for-seq-cst-stores-by-default-87376.tar.bz2 | |
Rebaseusers/ilovepi/spr/reapply-llvmriscv-enable-trailing-fences-for-seq-cst-stores-by-default-87376
Created using spr 1.3.4
619 files changed, 7252 insertions, 2630 deletions
diff --git a/bolt/include/bolt/Rewrite/DWARFRewriter.h b/bolt/include/bolt/Rewrite/DWARFRewriter.h index 4559ff5..c34fbd5 100644 --- a/bolt/include/bolt/Rewrite/DWARFRewriter.h +++ b/bolt/include/bolt/Rewrite/DWARFRewriter.h @@ -16,7 +16,6 @@ #include "llvm/ADT/StringRef.h" #include "llvm/CodeGen/DIE.h" #include "llvm/DWP/DWP.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCContext.h" #include "llvm/Support/ToolOutputFile.h" #include <cstdint> @@ -183,7 +182,7 @@ public: void updateDebugInfo(); /// Update stmt_list for CUs based on the new .debug_line \p Layout. - void updateLineTableOffsets(const MCAsmLayout &Layout); + void updateLineTableOffsets(const MCAssembler &Asm); uint64_t getDwoRangesBase(uint64_t DWOId) { return DwoRangesBase[DWOId]; } diff --git a/bolt/lib/Core/BinaryContext.cpp b/bolt/lib/Core/BinaryContext.cpp index 3bd715d..f28a0cd 100644 --- a/bolt/lib/Core/BinaryContext.cpp +++ b/bolt/lib/Core/BinaryContext.cpp @@ -20,7 +20,6 @@ #include "llvm/DebugInfo/DWARF/DWARFCompileUnit.h" #include "llvm/DebugInfo/DWARF/DWARFFormValue.h" #include "llvm/DebugInfo/DWARF/DWARFUnit.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCDisassembler/MCDisassembler.h" @@ -2416,8 +2415,7 @@ BinaryContext::calculateEmittedSize(BinaryFunction &BF, bool FixBranches) { MCAssembler &Assembler = static_cast<MCObjectStreamer *>(Streamer.get())->getAssembler(); - MCAsmLayout Layout(Assembler); - Assembler.layout(Layout); + Assembler.layout(); // Obtain fragment sizes. std::vector<uint64_t> FragmentSizes; diff --git a/bolt/lib/Rewrite/DWARFRewriter.cpp b/bolt/lib/Rewrite/DWARFRewriter.cpp index 110f53f..1f5c3a4 100644 --- a/bolt/lib/Rewrite/DWARFRewriter.cpp +++ b/bolt/lib/Rewrite/DWARFRewriter.cpp @@ -29,7 +29,6 @@ #include "llvm/DebugInfo/DWARF/DWARFTypeUnit.h" #include "llvm/DebugInfo/DWARF/DWARFUnit.h" #include "llvm/MC/MCAsmBackend.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCObjectWriter.h" #include "llvm/MC/MCStreamer.h" @@ -1352,7 +1351,7 @@ void DWARFRewriter::updateDWARFObjectAddressRanges( } } -void DWARFRewriter::updateLineTableOffsets(const MCAsmLayout &Layout) { +void DWARFRewriter::updateLineTableOffsets(const MCAssembler &Asm) { ErrorOr<BinarySection &> DbgInfoSection = BC.getUniqueSectionByName(".debug_info"); ErrorOr<BinarySection &> TypeInfoSection = @@ -1394,7 +1393,7 @@ void DWARFRewriter::updateLineTableOffsets(const MCAsmLayout &Layout) { continue; const uint64_t LineTableOffset = - Layout.getAssembler().getSymbolOffset(*Label); + Asm.getSymbolOffset(*Label); DebugLineOffsetMap[*StmtOffset] = LineTableOffset; assert(DbgInfoSection && ".debug_info section must exist"); LineTablePatchMap[CU.get()] = LineTableOffset; diff --git a/bolt/lib/Rewrite/RewriteInstance.cpp b/bolt/lib/Rewrite/RewriteInstance.cpp index c118f2c..8248c1c 100644 --- a/bolt/lib/Rewrite/RewriteInstance.cpp +++ b/bolt/lib/Rewrite/RewriteInstance.cpp @@ -40,7 +40,6 @@ #include "llvm/DebugInfo/DWARF/DWARFDebugFrame.h" #include "llvm/MC/MCAsmBackend.h" #include "llvm/MC/MCAsmInfo.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCDisassembler/MCDisassembler.h" #include "llvm/MC/MCObjectStreamer.h" #include "llvm/MC/MCStreamer.h" @@ -3508,9 +3507,8 @@ void RewriteInstance::emitAndLink() { updateOutputValues(*Linker); if (opts::UpdateDebugSections) { - MCAsmLayout FinalLayout( - static_cast<MCObjectStreamer *>(Streamer.get())->getAssembler()); - DebugInfoRewriter->updateLineTableOffsets(FinalLayout); + DebugInfoRewriter->updateLineTableOffsets( + static_cast<MCObjectStreamer &>(*Streamer).getAssembler()); } if (RuntimeLibrary *RtLibrary = BC->getRuntimeLibrary()) diff --git a/clang-tools-extra/clang-tidy/bugprone/SizeofExpressionCheck.cpp b/clang-tools-extra/clang-tidy/bugprone/SizeofExpressionCheck.cpp index c25ee42..d517e84 100644 --- a/clang-tools-extra/clang-tidy/bugprone/SizeofExpressionCheck.cpp +++ b/clang-tools-extra/clang-tidy/bugprone/SizeofExpressionCheck.cpp @@ -296,7 +296,7 @@ void SizeofExpressionCheck::check(const MatchFinder::MatchResult &Result) { } else if (const auto *E = Result.Nodes.getNodeAs<Expr>("sizeof-integer-call")) { diag(E->getBeginLoc(), "suspicious usage of 'sizeof()' on an expression " - "that results in an integer") + "of integer type") << E->getSourceRange(); } else if (const auto *E = Result.Nodes.getNodeAs<Expr>("sizeof-this")) { diag(E->getBeginLoc(), @@ -314,7 +314,7 @@ void SizeofExpressionCheck::check(const MatchFinder::MatchResult &Result) { << E->getSourceRange(); } else { diag(E->getBeginLoc(), "suspicious usage of 'sizeof()' on an expression " - "that results in a pointer") + "of pointer type") << E->getSourceRange(); } } else if (const auto *E = Result.Nodes.getNodeAs<BinaryOperator>( @@ -348,25 +348,28 @@ void SizeofExpressionCheck::check(const MatchFinder::MatchResult &Result) { } else if (ElementSize > CharUnits::Zero() && DenominatorSize > CharUnits::Zero() && ElementSize != DenominatorSize) { - diag(E->getOperatorLoc(), "suspicious usage of 'sizeof(...)/sizeof(...)';" - " numerator is not a multiple of denominator") + // FIXME: Apparently there are no testcases that cover this branch! + diag(E->getOperatorLoc(), + "suspicious usage of 'sizeof(array)/sizeof(...)';" + " denominator differs from the size of array elements") << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } else if (NumTy && DenomTy && NumTy == DenomTy) { - // FIXME: This message is wrong, it should not refer to sizeof "pointer" - // usage (and by the way, it would be to clarify all the messages). diag(E->getOperatorLoc(), - "suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)'") + "suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions " + "have the same type") << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } else if (!WarnOnSizeOfPointer) { // When 'WarnOnSizeOfPointer' is enabled, these messages become redundant: if (PointedTy && DenomTy && PointedTy == DenomTy) { diag(E->getOperatorLoc(), - "suspicious usage of sizeof pointer 'sizeof(T*)/sizeof(T)'") + "suspicious usage of 'sizeof(...)/sizeof(...)'; size of pointer " + "is divided by size of pointed type") << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } else if (NumTy && DenomTy && NumTy->isPointerType() && DenomTy->isPointerType()) { diag(E->getOperatorLoc(), - "suspicious usage of sizeof pointer 'sizeof(P*)/sizeof(Q*)'") + "suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions " + "have pointer types") << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } } diff --git a/clang-tools-extra/docs/ReleaseNotes.rst b/clang-tools-extra/docs/ReleaseNotes.rst index 4d2cd0a..fe81973 100644 --- a/clang-tools-extra/docs/ReleaseNotes.rst +++ b/clang-tools-extra/docs/ReleaseNotes.rst @@ -241,10 +241,10 @@ Changes in existing checks false positives resulting from use of optionals in unevaluated context. - Improved :doc:`bugprone-sizeof-expression - <clang-tidy/checks/bugprone/sizeof-expression>` check by eliminating some - false positives and adding a new (off-by-default) option - `WarnOnSizeOfPointer` that reports all ``sizeof(pointer)`` expressions - (except for a few that are idiomatic). + <clang-tidy/checks/bugprone/sizeof-expression>` check by clarifying the + diagnostics, eliminating some false positives and adding a new + (off-by-default) option `WarnOnSizeOfPointer` that reports all + ``sizeof(pointer)`` expressions (except for a few that are idiomatic). - Improved :doc:`bugprone-suspicious-include <clang-tidy/checks/bugprone/suspicious-include>` check by replacing the local diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-2.c b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-2.c index aef930f..b898071 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-2.c +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-2.c @@ -34,24 +34,24 @@ int Test5() { int sum = 0; sum += sizeof(&S); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(__typeof(&S)); sum += sizeof(&TS); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(__typeof(&TS)); sum += sizeof(STRKWD MyStruct*); sum += sizeof(__typeof(STRKWD MyStruct*)); sum += sizeof(TypedefStruct*); sum += sizeof(__typeof(TypedefStruct*)); sum += sizeof(PTTS); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PMyStruct); sum += sizeof(PS); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PS2); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&A10); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type #ifdef __cplusplus MyStruct &rS = S; diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-any-pointer.cpp b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-any-pointer.cpp index bfb2ec3..4c65d25 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-any-pointer.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-any-pointer.cpp @@ -38,17 +38,17 @@ int Test1(const char* ptr) { sum += sizeof(sum, LEN); // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: suspicious usage of 'sizeof(..., ...)' sum += sizeof(AsBool()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(AsInt()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(AsEnum()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(AsEnumClass()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(M{}.AsInt()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(M{}.AsEnum()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(sizeof(X)); // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof(sizeof(...))' sum += sizeof(LEN + sizeof(X)); @@ -62,7 +62,7 @@ int Test1(const char* ptr) { sum += sizeof(LEN + - + -sizeof(X)); // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof(sizeof(...))' sum += sizeof(char) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type sum += sizeof(A) / sizeof(S); // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; numerator is not a multiple of denominator sum += sizeof(char) / sizeof(int); @@ -72,21 +72,21 @@ int Test1(const char* ptr) { sum += sizeof(B[0]) / sizeof(A); // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; numerator is not a multiple of denominator sum += sizeof(ptr) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(ptr) / sizeof(ptr[0]); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(ptr) / sizeof(char*); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(ptr) / sizeof(void*); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(ptr) / sizeof(const void volatile*); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(ptr) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(int) * sizeof(char); // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious 'sizeof' by 'sizeof' multiplication sum += sizeof(ptr) * sizeof(ptr[0]); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type // CHECK-MESSAGES: :[[@LINE-2]]:22: warning: suspicious 'sizeof' by 'sizeof' multiplication sum += sizeof(int) * (2 * sizeof(char)); // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious 'sizeof' by 'sizeof' multiplication @@ -127,54 +127,54 @@ int Test5() { int sum = 0; sum += sizeof(&S.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&kGlocalMyStruct.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&kGlocalMyStructPtr->arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(S.arr + 0); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(+ S.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof((int*)S.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(S.ptr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(kGlocalMyStruct.ptr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(kGlocalMyStructPtr->ptr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&kGlocalMyStruct); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&S); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(MyStruct*); sum += sizeof(PMyStruct); sum += sizeof(PS); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PS2); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&A10); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PtrArray) / sizeof(PtrArray[1]); - // CHECK-MESSAGES: :[[@LINE-1]]:29: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:29: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(A10) / sizeof(PtrArray[0]); sum += sizeof(PC) / sizeof(PtrArray[0]); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer - // CHECK-MESSAGES: :[[@LINE-2]]:21: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type + // CHECK-MESSAGES: :[[@LINE-2]]:21: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type sum += sizeof(ArrayC) / sizeof(PtrArray[0]); // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; numerator is not a multiple of denominator sum += sizeof(PChar); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PInt); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PPInt); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PPMyStruct); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type return sum; } @@ -200,9 +200,9 @@ void GenericFunctionTest() { // reported by the `sizeof(pointer)` checks, but this causes some false // positives, so it would be good to create an exception for them. some_generic_function(&IntPP, sizeof(IntP)); - // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: suspicious usage of 'sizeof()' on an expression of pointer type some_generic_function(&ClassPP, sizeof(ClassP)); - // CHECK-MESSAGES: :[[@LINE-1]]:35: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:35: warning: suspicious usage of 'sizeof()' on an expression of pointer type } int ValidExpressions() { @@ -222,7 +222,7 @@ int ValidExpressions() { sum += sizeof(A[sizeof(A) / sizeof(int)]); // Here the outer sizeof is reported, but the inner ones are accepted: sum += sizeof(&A[sizeof(A) / sizeof(int)]); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(sizeof(0)); // Special case: sizeof size_t. sum += sizeof(void*); sum += sizeof(void const *); diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-warn-on-sizeof-pointer-to-aggregate.cpp b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-warn-on-sizeof-pointer-to-aggregate.cpp index 5ef2c46..5c6b728 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-warn-on-sizeof-pointer-to-aggregate.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression-warn-on-sizeof-pointer-to-aggregate.cpp @@ -68,7 +68,7 @@ int Test5() { sum += sizeof(A10) / sizeof(PtrArray[0]); // No warning. sum += sizeof(PC) / sizeof(PtrArray[0]); - // CHECK-MESSAGES: :[[@LINE-1]]:21: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:21: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type sum += sizeof(ArrayC) / sizeof(PtrArray[0]); // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; numerator is not a multiple of denominator diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression.cpp b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression.cpp index 064f31c..671fd83 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/sizeof-expression.cpp @@ -48,13 +48,13 @@ bool TestTrait2() { template <class T> bool TestTrait3() { return sizeof(ReturnOverload(0)) == sizeof(T{}); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type } template <class T> bool TestTrait4() { return sizeof(ReturnTemplate(0)) == sizeof(T{}); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type } bool TestTemplates() { @@ -79,17 +79,17 @@ int Test1(const char* ptr) { sum += sizeof(sum, LEN); // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: suspicious usage of 'sizeof(..., ...)' sum += sizeof(AsBool()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(AsInt()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(AsEnum()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(AsEnumClass()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(M{}.AsInt()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(M{}.AsEnum()); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in an integer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of integer type sum += sizeof(sizeof(X)); // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof(sizeof(...))' sum += sizeof(LEN + sizeof(X)); @@ -103,7 +103,7 @@ int Test1(const char* ptr) { sum += sizeof(LEN + - + -sizeof(X)); // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof(sizeof(...))' sum += sizeof(char) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type sum += sizeof(A) / sizeof(S); // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; numerator is not a multiple of denominator sum += sizeof(char) / sizeof(int); @@ -113,17 +113,17 @@ int Test1(const char* ptr) { sum += sizeof(B[0]) / sizeof(A); // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; numerator is not a multiple of denominator sum += sizeof(ptr) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of sizeof pointer 'sizeof(T*)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; size of pointer is divided by size of pointed type sum += sizeof(ptr) / sizeof(ptr[0]); - // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of sizeof pointer 'sizeof(T*)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; size of pointer is divided by size of pointed type sum += sizeof(ptr) / sizeof(char*); - // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of sizeof pointer 'sizeof(P*)/sizeof(Q*)' + // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have pointer types sum += sizeof(ptr) / sizeof(void*); - // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of sizeof pointer 'sizeof(P*)/sizeof(Q*)' + // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have pointer types sum += sizeof(ptr) / sizeof(const void volatile*); - // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of sizeof pointer 'sizeof(P*)/sizeof(Q*)' + // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have pointer types sum += sizeof(ptr) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of sizeof pointer 'sizeof(T*)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; size of pointer is divided by size of pointed type sum += sizeof(int) * sizeof(char); // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: suspicious 'sizeof' by 'sizeof' multiplication sum += sizeof(ptr) * sizeof(ptr[0]); @@ -156,11 +156,11 @@ int CE4 = sizeof sizeof(MyConstChar); int Test2(MyConstChar* A) { int sum = 0; sum += sizeof(MyConstChar) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:30: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:30: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type sum += sizeof(MyConstChar) / sizeof(MyChar); - // CHECK-MESSAGES: :[[@LINE-1]]:30: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:30: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type sum += sizeof(A[0]) / sizeof(char); - // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type return sum; } @@ -169,7 +169,7 @@ int Foo() { int A[T]; return sizeof(T); } // CHECK-MESSAGES: :[[@LINE-1]]:30: warning: suspicious usage of 'sizeof(K)' template <typename T> int Bar() { T A[5]; return sizeof(A[0]) / sizeof(T); } -// CHECK-MESSAGES: :[[@LINE-1]]:41: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' +// CHECK-MESSAGES: :[[@LINE-1]]:41: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type int Test3() { return Foo<42>() + Bar<char>(); } static const char* kABC = "abc"; @@ -211,43 +211,43 @@ int Test5() { int sum = 0; sum += sizeof(&S.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&kGlocalMyStruct.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&kGlocalMyStructPtr->arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(S.arr + 0); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(+ S.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof((int*)S.arr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(S.ptr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(kGlocalMyStruct.ptr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(kGlocalMyStructPtr->ptr); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&kGlocalMyStruct); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&S); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(MyStruct*); sum += sizeof(PMyStruct); sum += sizeof(PS); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PS2); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(&A10); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(PtrArray) / sizeof(PtrArray[1]); - // CHECK-MESSAGES: :[[@LINE-1]]:29: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:29: warning: suspicious usage of 'sizeof()' on an expression of pointer type sum += sizeof(A10) / sizeof(PtrArray[0]); sum += sizeof(PC) / sizeof(PtrArray[0]); - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer - // CHECK-MESSAGES: :[[@LINE-2]]:21: warning: suspicious usage of sizeof pointer 'sizeof(T)/sizeof(T)' + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: suspicious usage of 'sizeof()' on an expression of pointer type + // CHECK-MESSAGES: :[[@LINE-2]]:21: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; both expressions have the same type sum += sizeof(ArrayC) / sizeof(PtrArray[0]); // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: suspicious usage of 'sizeof(...)/sizeof(...)'; numerator is not a multiple of denominator @@ -321,7 +321,7 @@ void GenericFunctionTest() { // NOTE: `sizeof(IntP)` is only reported with `WarnOnSizeOfPointer=true`. some_generic_function(&IntPP, sizeof(IntP)); some_generic_function(&ClassPP, sizeof(ClassP)); - // CHECK-MESSAGES: :[[@LINE-1]]:35: warning: suspicious usage of 'sizeof()' on an expression that results in a pointer + // CHECK-MESSAGES: :[[@LINE-1]]:35: warning: suspicious usage of 'sizeof()' on an expression of pointer type } int ValidExpressions() { diff --git a/clang/docs/ClangFormatStyleOptions.rst b/clang/docs/ClangFormatStyleOptions.rst index 080cba9..6c2e6da 100644 --- a/clang/docs/ClangFormatStyleOptions.rst +++ b/clang/docs/ClangFormatStyleOptions.rst @@ -6242,6 +6242,7 @@ the configuration (without a prefix: ``Auto``). # Example of usage: SpacesInParens: Custom SpacesInParensOptions: + ExceptDoubleParentheses: false InConditionalStatements: true InEmptyParentheses: true @@ -6254,9 +6255,22 @@ the configuration (without a prefix: ``Auto``). # Should be declared this way: SpacesInParens: Custom SpacesInParensOptions: + ExceptDoubleParentheses: false InConditionalStatements: true Other: true + * ``bool ExceptDoubleParentheses`` Override any of the following options to prevent addition of space + when both opening and closing parentheses use multiple parentheses. + + .. code-block:: c++ + + true: + __attribute__(( noreturn )) + __decltype__(( x )) + if (( a = b )) + false: + Uses the applicable option. + * ``bool InConditionalStatements`` Put a space in parentheses only inside conditional statements (``for/if/while/switch...``). @@ -6270,8 +6284,9 @@ the configuration (without a prefix: ``Auto``). .. code-block:: c++ - true: false: - x = ( int32 )y vs. x = (int32)y + true: false: + x = ( int32 )y vs. x = (int32)y + y = (( int (*)(int) )foo)(x); y = ((int (*)(int))foo)(x); * ``bool InEmptyParentheses`` Insert a space in empty parentheses, i.e. ``()``. @@ -6289,8 +6304,8 @@ the configuration (without a prefix: ``Auto``). .. code-block:: c++ - true: false: - t f( Deleted & ) & = delete; vs. t f(Deleted &) & = delete; + true: false: + t f( Deleted & ) & = delete; vs. t f(Deleted &) & = delete; .. _SpacesInParentheses: diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst index 90c2469..181d100 100644 --- a/clang/docs/ReleaseNotes.rst +++ b/clang/docs/ReleaseNotes.rst @@ -320,6 +320,12 @@ Resolutions to C++ Defect Reports C Language Changes ------------------ +C2y Feature Support +^^^^^^^^^^^^^^^^^^^ +- Clang now enables C2y mode with ``-std=c2y``. This sets ``__STDC_VERSION__`` + to ``202400L`` so that it's greater than the value for C23. The value of this + macro is subject to change in the future. + C23 Feature Support ^^^^^^^^^^^^^^^^^^^ - No longer diagnose use of binary literals as an extension in C23 mode. Fixes @@ -403,6 +409,11 @@ Non-comprehensive list of changes in this release - ``#pragma GCC diagnostic warning "-Wfoo"`` can now downgrade ``-Werror=foo`` errors and certain default-to-error ``-W`` diagnostics to warnings. +- Support importing C++20 modules in clang-repl. + +- Added support for ``TypeLoc::dump()`` for easier debugging, and improved + textual and JSON dumping for various ``TypeLoc``-related nodes. + New Compiler Flags ------------------ - ``-fsanitize=implicit-bitfield-conversion`` checks implicit truncation and @@ -645,6 +656,8 @@ Improvements to Clang's diagnostics that will be destroyed at the end of the full expression. Fixes #GH54492. +- Clang now shows implicit deduction guides when diagnosing overload resolution failure. #GH92393. + Improvements to Clang's time-trace ---------------------------------- @@ -1130,6 +1143,9 @@ clang-format - Adds ``LeftWithLastLine`` suboption to ``AlignEscapedNewlines``. - Adds ``KeepEmptyLines`` option to deprecate ``KeepEmptyLinesAtEOF`` and ``KeepEmptyLinesAtTheStartOfBlocks``. +- Add ``ExceptDoubleParentheses`` sub-option for ``SpacesInParensOptions`` + to override addition of spaces between multiple, non-redundant parentheses + similar to the rules used for ``RemoveParentheses``. libclang -------- @@ -1158,11 +1174,6 @@ Crash and bug fixes Improvements ^^^^^^^^^^^^ -- Support importing C++20 modules in clang-repl. - -- Added support for ``TypeLoc::dump()`` for easier debugging, and improved - textual and JSON dumping for various ``TypeLoc``-related nodes. - Moved checkers ^^^^^^^^^^^^^^ diff --git a/clang/docs/UsersManual.rst b/clang/docs/UsersManual.rst index d273102..087c75b 100644 --- a/clang/docs/UsersManual.rst +++ b/clang/docs/UsersManual.rst @@ -3458,9 +3458,9 @@ Differences between various standard modes clang supports the -std option, which changes what language mode clang uses. The supported modes for C are c89, gnu89, c94, c99, gnu99, c11, gnu11, c17, -gnu17, c23, gnu23, and various aliases for those modes. If no -std option is -specified, clang defaults to gnu17 mode. Many C99 and C11 features are -supported in earlier modes as a conforming extension, with a warning. Use +gnu17, c23, gnu23, c2y, gnu2y, and various aliases for those modes. If no -std +option is specified, clang defaults to gnu17 mode. Many C99 and C11 features +are supported in earlier modes as a conforming extension, with a warning. Use ``-pedantic-errors`` to request an error if a feature from a later standard revision is used in an earlier mode. @@ -3523,6 +3523,10 @@ Differences between ``*17`` and ``*23`` modes: - ``[[]]`` attributes are supported by default in ``*23`` mode, and as an extension in ``*17`` and earlier modes. +Differences between ``*23`` and ``*2y`` modes: + +- ``__STDC_VERSION__`` is defined to ``202400L`` rather than ``202311L``. + GCC extensions not implemented yet ---------------------------------- diff --git a/clang/docs/tools/clang-formatted-files.txt b/clang/docs/tools/clang-formatted-files.txt index 4866bd4..a8ee8f1 100644 --- a/clang/docs/tools/clang-formatted-files.txt +++ b/clang/docs/tools/clang-formatted-files.txt @@ -5357,7 +5357,6 @@ llvm/include/llvm/MC/MCAsmInfoELF.h llvm/include/llvm/MC/MCAsmInfoGOFF.h llvm/include/llvm/MC/MCAsmInfoWasm.h llvm/include/llvm/MC/MCAsmInfoXCOFF.h -llvm/include/llvm/MC/MCAsmLayout.h llvm/include/llvm/MC/MCCodeView.h llvm/include/llvm/MC/MCContext.h llvm/include/llvm/MC/MCFixedLenDisassembler.h diff --git a/clang/include/clang/AST/Redeclarable.h b/clang/include/clang/AST/Redeclarable.h index 091bb88..74ccd74 100644 --- a/clang/include/clang/AST/Redeclarable.h +++ b/clang/include/clang/AST/Redeclarable.h @@ -281,10 +281,10 @@ public: return tmp; } - friend bool operator==(redecl_iterator x, redecl_iterator y) { + friend bool operator==(const redecl_iterator &x, const redecl_iterator &y) { return x.Current == y.Current; } - friend bool operator!=(redecl_iterator x, redecl_iterator y) { + friend bool operator!=(const redecl_iterator &x, const redecl_iterator &y) { return x.Current != y.Current; } }; diff --git a/clang/include/clang/AST/TemplateName.h b/clang/include/clang/AST/TemplateName.h index 24a7fde..e3b7dd2 100644 --- a/clang/include/clang/AST/TemplateName.h +++ b/clang/include/clang/AST/TemplateName.h @@ -198,7 +198,8 @@ public: /// /// Here, "apply" is treated as a template name within the typename /// specifier in the typedef. "apply" is a nested template, and can -/// only be understood in the context of +/// only be understood in the context of a template instantiation, +/// hence is represented as a dependent template name. class TemplateName { // NameDecl is either a TemplateDecl or a UsingShadowDecl depending on the // NameKind. diff --git a/clang/include/clang/Basic/DiagnosticGroups.td b/clang/include/clang/Basic/DiagnosticGroups.td index 74deb92..9431eea 100644 --- a/clang/include/clang/Basic/DiagnosticGroups.td +++ b/clang/include/clang/Basic/DiagnosticGroups.td @@ -299,6 +299,10 @@ def CPre23CompatPedantic : DiagGroup<"pre-c23-compat-pedantic", def : DiagGroup<"pre-c2x-compat", [CPre23Compat]>; def : DiagGroup<"pre-c2x-compat-pedantic", [CPre23CompatPedantic]>; +def CPre2yCompat : DiagGroup<"pre-c2y-compat">; +def CPre2yCompatPedantic : DiagGroup<"pre-c2y-compat-pedantic", + [CPre2yCompat]>; + // Warnings for C++ code which is not compatible with previous C++ standards. def CXXPre14Compat : DiagGroup<"pre-c++14-compat">; def : DiagGroup<"c++98-c++11-compat", [CXXPre14Compat]>; @@ -1197,6 +1201,9 @@ def C23 : DiagGroup<"c23-extensions">; def : DiagGroup<"c2x-extensions", [C23]>; +// A warning group for warnings about using C2y features as extensions. +def C2y : DiagGroup<"c2y-extensions">; + // Previously supported warning group which is no longer pertinent as binary // literals are a C++14 and C23 extension now instead of a GNU extension. def GNUBinaryLiteral : DiagGroup<"gnu-binary-literal">; diff --git a/clang/include/clang/Basic/DiagnosticParseKinds.td b/clang/include/clang/Basic/DiagnosticParseKinds.td index 1160b0f..12aab09 100644 --- a/clang/include/clang/Basic/DiagnosticParseKinds.td +++ b/clang/include/clang/Basic/DiagnosticParseKinds.td @@ -157,9 +157,12 @@ def err_duplicate_default_assoc : Error< "duplicate default generic association">; def note_previous_default_assoc : Note< "previous default generic association is here">; -def ext_generic_with_type_arg : Extension< - "passing a type argument as the first operand to '_Generic' is a Clang " - "extension">, InGroup<DiagGroup<"generic-type-extension">>; +def ext_c2y_generic_with_type_arg : Extension< + "passing a type argument as the first operand to '_Generic' is a C2y " + "extension">, InGroup<C2y>; +def warn_c2y_compat_generic_with_type_arg : Warning< + "passing a type argument as the first operand to '_Generic' is incompatible " + "with C standards before C2y">, InGroup<CPre2yCompat>, DefaultIgnore; def ext_c99_feature : Extension< "'%0' is a C99 extension">, InGroup<C99>; diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td b/clang/include/clang/Basic/DiagnosticSemaKinds.td index 64f7935d..3df64b2e 100644 --- a/clang/include/clang/Basic/DiagnosticSemaKinds.td +++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td @@ -2419,6 +2419,7 @@ def err_selected_explicit_constructor : Error< "chosen constructor is explicit in copy-initialization">; def note_explicit_ctor_deduction_guide_here : Note< "explicit %select{constructor|deduction guide}0 declared here">; +def note_implicit_deduction_guide : Note<"implicit deduction guide declared as '%0'">; // C++11 auto def warn_cxx98_compat_auto_type_specifier : Warning< @@ -3546,6 +3547,12 @@ def warn_alignment_not_power_of_two : Warning< InGroup<DiagGroup<"non-power-of-two-alignment">>; def err_alignment_dependent_typedef_name : Error< "requested alignment is dependent but declaration is not dependent">; +def ext_c2y_alignof_incomplete_array : Extension< + "'alignof' on an incomplete array type is a C2y extension">, + InGroup<C2y>; +def warn_c2y_compat_alignof_incomplete_array : Warning< + "'alignof' on an incomplete array type is incompatible with C standards " + "before C2y">, InGroup<CPre2yCompat>, DefaultIgnore; def warn_alignment_builtin_useless : Warning< "%select{aligning a value|the result of checking whether a value is aligned}0" @@ -7663,8 +7670,12 @@ def ext_gnu_ptr_func_arith : Extension< InGroup<GNUPointerArith>; def err_readonly_message_assignment : Error< "assigning to 'readonly' return result of an Objective-C message not allowed">; -def ext_increment_complex : Extension< - "'%select{--|++}0' on an object of complex type is a Clang extension">; +def ext_c2y_increment_complex : Extension< + "'%select{--|++}0' on an object of complex type is a C2y extension">, + InGroup<C2y>; +def warn_c2y_compat_increment_complex : Warning< + "'%select{--|++}0' on an object of complex type is incompatible with C " + "standards before C2y">, InGroup<CPre2yCompat>, DefaultIgnore; def ext_integer_complement_complex : Extension< "ISO C does not support '~' for complex conjugation of %0">; def err_nosetter_property_assignment : Error< diff --git a/clang/include/clang/Basic/LangOptions.def b/clang/include/clang/Basic/LangOptions.def index 6dd6b56..491759e 100644 --- a/clang/include/clang/Basic/LangOptions.def +++ b/clang/include/clang/Basic/LangOptions.def @@ -87,6 +87,7 @@ LANGOPT(C99 , 1, 0, "C99") LANGOPT(C11 , 1, 0, "C11") LANGOPT(C17 , 1, 0, "C17") LANGOPT(C23 , 1, 0, "C23") +LANGOPT(C2y , 1, 0, "C2y") LANGOPT(MSVCCompat , 1, 0, "Microsoft Visual C++ full compatibility mode") LANGOPT(Kernel , 1, 0, "Kernel mode") LANGOPT(MicrosoftExt , 1, 0, "Microsoft C++ extensions") diff --git a/clang/include/clang/Basic/LangStandard.h b/clang/include/clang/Basic/LangStandard.h index 8e25afc..f79b4aa 100644 --- a/clang/include/clang/Basic/LangStandard.h +++ b/clang/include/clang/Basic/LangStandard.h @@ -52,18 +52,19 @@ enum LangFeatures { C11 = (1 << 2), C17 = (1 << 3), C23 = (1 << 4), - CPlusPlus = (1 << 5), - CPlusPlus11 = (1 << 6), - CPlusPlus14 = (1 << 7), - CPlusPlus17 = (1 << 8), - CPlusPlus20 = (1 << 9), - CPlusPlus23 = (1 << 10), - CPlusPlus26 = (1 << 11), - Digraphs = (1 << 12), - GNUMode = (1 << 13), - HexFloat = (1 << 14), - OpenCL = (1 << 15), - HLSL = (1 << 16) + C2y = (1 << 5), + CPlusPlus = (1 << 6), + CPlusPlus11 = (1 << 7), + CPlusPlus14 = (1 << 8), + CPlusPlus17 = (1 << 9), + CPlusPlus20 = (1 << 10), + CPlusPlus23 = (1 << 11), + CPlusPlus26 = (1 << 12), + Digraphs = (1 << 13), + GNUMode = (1 << 14), + HexFloat = (1 << 15), + OpenCL = (1 << 16), + HLSL = (1 << 17) }; /// LangStandard - Information about the properties of a particular language @@ -106,6 +107,9 @@ public: /// isC23 - Language is a superset of C23. bool isC23() const { return Flags & C23; } + /// isC2y - Language is a superset of C2y. + bool isC2y() const { return Flags & C2y; } + /// isCPlusPlus - Language is a C++ variant. bool isCPlusPlus() const { return Flags & CPlusPlus; } diff --git a/clang/include/clang/Basic/LangStandards.def b/clang/include/clang/Basic/LangStandards.def index b6192e4..f0c2593 100644 --- a/clang/include/clang/Basic/LangStandards.def +++ b/clang/include/clang/Basic/LangStandards.def @@ -99,6 +99,15 @@ LANGSTANDARD_ALIAS_DEPR(gnu23, "gnu2x") // FIXME: Add the alias for iso9899:202* once we know the year ISO publishes // the document (expected to be 2024). +// C2y modes +LANGSTANDARD(c2y, "c2y", + C, "Working Draft for ISO C2y", + LineComment | C99 | C11 | C17 | C23 | C2y | Digraphs | HexFloat) +LANGSTANDARD(gnu2y, "gnu2y", + C, "Working Draft for ISO C2y with GNU extensions", + LineComment | C99 | C11 | C17 | C23 | C2y | Digraphs | GNUMode | HexFloat) + + // C++ modes LANGSTANDARD(cxx98, "c++98", CXX, "ISO C++ 1998 with amendments", diff --git a/clang/include/clang/Driver/ToolChain.h b/clang/include/clang/Driver/ToolChain.h index 1f93bd6..ece1384 100644 --- a/clang/include/clang/Driver/ToolChain.h +++ b/clang/include/clang/Driver/ToolChain.h @@ -526,6 +526,9 @@ public: // Returns target specific standard library path if it exists. std::optional<std::string> getStdlibPath() const; + // Returns target specific standard library include path if it exists. + std::optional<std::string> getStdlibIncludePath() const; + // Returns <ResourceDir>/lib/<OSName>/<arch> or <ResourceDir>/lib/<triple>. // This is used by runtimes (such as OpenMP) to find arch-specific libraries. virtual path_list getArchSpecificLibPaths() const; diff --git a/clang/include/clang/Format/Format.h b/clang/include/clang/Format/Format.h index efc2e450..c454ab2 100644 --- a/clang/include/clang/Format/Format.h +++ b/clang/include/clang/Format/Format.h @@ -4679,10 +4679,22 @@ struct FormatStyle { /// # Should be declared this way: /// SpacesInParens: Custom /// SpacesInParensOptions: + /// ExceptDoubleParentheses: false /// InConditionalStatements: true /// Other: true /// \endcode struct SpacesInParensCustom { + /// Override any of the following options to prevent addition of space + /// when both opening and closing parentheses use multiple parentheses. + /// \code + /// true: + /// __attribute__(( noreturn )) + /// __decltype__(( x )) + /// if (( a = b )) + /// \endcode + /// false: + /// Uses the applicable option. + bool ExceptDoubleParentheses; /// Put a space in parentheses only inside conditional statements /// (``for/if/while/switch...``). /// \code @@ -4693,8 +4705,9 @@ struct FormatStyle { bool InConditionalStatements; /// Put a space in C style casts. /// \code - /// true: false: - /// x = ( int32 )y vs. x = (int32)y + /// true: false: + /// x = ( int32 )y vs. x = (int32)y + /// y = (( int (*)(int) )foo)(x); y = ((int (*)(int))foo)(x); /// \endcode bool InCStyleCasts; /// Insert a space in empty parentheses, i.e. ``()``. @@ -4710,23 +4723,26 @@ struct FormatStyle { bool InEmptyParentheses; /// Put a space in parentheses not covered by preceding options. /// \code - /// true: false: - /// t f( Deleted & ) & = delete; vs. t f(Deleted &) & = delete; + /// true: false: + /// t f( Deleted & ) & = delete; vs. t f(Deleted &) & = delete; /// \endcode bool Other; SpacesInParensCustom() - : InConditionalStatements(false), InCStyleCasts(false), - InEmptyParentheses(false), Other(false) {} + : ExceptDoubleParentheses(false), InConditionalStatements(false), + InCStyleCasts(false), InEmptyParentheses(false), Other(false) {} - SpacesInParensCustom(bool InConditionalStatements, bool InCStyleCasts, + SpacesInParensCustom(bool ExceptDoubleParentheses, + bool InConditionalStatements, bool InCStyleCasts, bool InEmptyParentheses, bool Other) - : InConditionalStatements(InConditionalStatements), + : ExceptDoubleParentheses(ExceptDoubleParentheses), + InConditionalStatements(InConditionalStatements), InCStyleCasts(InCStyleCasts), InEmptyParentheses(InEmptyParentheses), Other(Other) {} bool operator==(const SpacesInParensCustom &R) const { - return InConditionalStatements == R.InConditionalStatements && + return ExceptDoubleParentheses == R.ExceptDoubleParentheses && + InConditionalStatements == R.InConditionalStatements && InCStyleCasts == R.InCStyleCasts && InEmptyParentheses == R.InEmptyParentheses && Other == R.Other; } @@ -4744,6 +4760,7 @@ struct FormatStyle { /// # Example of usage: /// SpacesInParens: Custom /// SpacesInParensOptions: + /// ExceptDoubleParentheses: false /// InConditionalStatements: true /// InEmptyParentheses: true /// \endcode diff --git a/clang/lib/AST/Interp/Compiler.cpp b/clang/lib/AST/Interp/Compiler.cpp index 5ae36da..9ca71e0 100644 --- a/clang/lib/AST/Interp/Compiler.cpp +++ b/clang/lib/AST/Interp/Compiler.cpp @@ -154,6 +154,19 @@ private: CaseMap OldCaseLabels; }; +template <class Emitter> class StmtExprScope final { +public: + StmtExprScope(Compiler<Emitter> *Ctx) : Ctx(Ctx), OldFlag(Ctx->InStmtExpr) { + Ctx->InStmtExpr = true; + } + + ~StmtExprScope() { Ctx->InStmtExpr = OldFlag; } + +private: + Compiler<Emitter> *Ctx; + bool OldFlag; +}; + } // namespace interp } // namespace clang @@ -3028,6 +3041,7 @@ bool Compiler<Emitter>::VisitCXXStdInitializerListExpr( template <class Emitter> bool Compiler<Emitter>::VisitStmtExpr(const StmtExpr *E) { BlockScope<Emitter> BS(this); + StmtExprScope<Emitter> SS(this); const CompoundStmt *CS = E->getSubStmt(); const Stmt *Result = CS->getStmtExprResult(); @@ -3460,7 +3474,7 @@ bool Compiler<Emitter>::visitDecl(const VarDecl *VD, bool ConstantContext) { } // Create and initialize the variable. - if (!this->visitVarDecl(VD)) + if (!this->visitVarDecl(VD, /*Toplevel=*/true)) return false; // Get a pointer to the variable @@ -3507,7 +3521,7 @@ bool Compiler<Emitter>::visitDecl(const VarDecl *VD, bool ConstantContext) { } template <class Emitter> -VarCreationState Compiler<Emitter>::visitVarDecl(const VarDecl *VD) { +VarCreationState Compiler<Emitter>::visitVarDecl(const VarDecl *VD, bool Toplevel) { // We don't know what to do with these, so just return false. if (VD->getType().isNull()) return false; @@ -3521,7 +3535,7 @@ VarCreationState Compiler<Emitter>::visitVarDecl(const VarDecl *VD) { std::optional<PrimType> VarT = classify(VD->getType()); auto checkDecl = [&]() -> bool { - bool NeedsOp = VD->isLocalVarDecl() && VD->isStaticLocal(); + bool NeedsOp = !Toplevel && VD->isLocalVarDecl() && VD->isStaticLocal(); return !NeedsOp || this->emitCheckDecl(VD, VD); }; @@ -4019,7 +4033,7 @@ template <class Emitter> bool Compiler<Emitter>::visitLoopBody(const Stmt *S) { return true; if (const auto *CS = dyn_cast<CompoundStmt>(S)) { - for (auto *InnerStmt : CS->body()) + for (const auto *InnerStmt : CS->body()) if (!visitStmt(InnerStmt)) return false; return true; @@ -4031,7 +4045,7 @@ template <class Emitter> bool Compiler<Emitter>::visitLoopBody(const Stmt *S) { template <class Emitter> bool Compiler<Emitter>::visitCompoundStmt(const CompoundStmt *S) { BlockScope<Emitter> Scope(this); - for (auto *InnerStmt : S->body()) + for (const auto *InnerStmt : S->body()) if (!visitStmt(InnerStmt)) return false; return true; @@ -4056,6 +4070,9 @@ bool Compiler<Emitter>::visitDeclStmt(const DeclStmt *DS) { template <class Emitter> bool Compiler<Emitter>::visitReturnStmt(const ReturnStmt *RS) { + if (this->InStmtExpr) + return this->emitUnsupported(RS); + if (const Expr *RE = RS->getRetValue()) { ExprScope<Emitter> RetScope(this); if (ReturnType) { @@ -4139,6 +4156,7 @@ bool Compiler<Emitter>::visitWhileStmt(const WhileStmt *S) { LabelTy EndLabel = this->getLabel(); // Label after the loop. LoopScope<Emitter> LS(this, EndLabel, CondLabel); + this->fallthrough(CondLabel); this->emitLabel(CondLabel); if (const DeclStmt *CondDecl = S->getConditionVariableDeclStmt()) @@ -4174,12 +4192,14 @@ template <class Emitter> bool Compiler<Emitter>::visitDoStmt(const DoStmt *S) { LoopScope<Emitter> LS(this, EndLabel, CondLabel); LocalScope<Emitter> Scope(this); + this->fallthrough(StartLabel); this->emitLabel(StartLabel); { DestructorScope<Emitter> DS(Scope); if (!this->visitLoopBody(Body)) return false; + this->fallthrough(CondLabel); this->emitLabel(CondLabel); if (!this->visitBool(Cond)) return false; @@ -4187,6 +4207,7 @@ template <class Emitter> bool Compiler<Emitter>::visitDoStmt(const DoStmt *S) { if (!this->jumpTrue(StartLabel)) return false; + this->fallthrough(EndLabel); this->emitLabel(EndLabel); return true; } @@ -4207,6 +4228,7 @@ bool Compiler<Emitter>::visitForStmt(const ForStmt *S) { if (Init && !this->visitStmt(Init)) return false; + this->fallthrough(CondLabel); this->emitLabel(CondLabel); if (const DeclStmt *CondDecl = S->getConditionVariableDeclStmt()) @@ -4225,6 +4247,7 @@ bool Compiler<Emitter>::visitForStmt(const ForStmt *S) { if (Body && !this->visitLoopBody(Body)) return false; + this->fallthrough(IncLabel); this->emitLabel(IncLabel); if (Inc && !this->discard(Inc)) return false; @@ -4232,6 +4255,7 @@ bool Compiler<Emitter>::visitForStmt(const ForStmt *S) { if (!this->jump(CondLabel)) return false; + this->fallthrough(EndLabel); this->emitLabel(EndLabel); return true; } @@ -4263,6 +4287,7 @@ bool Compiler<Emitter>::visitCXXForRangeStmt(const CXXForRangeStmt *S) { return false; // Now the condition as well as the loop variable assignment. + this->fallthrough(CondLabel); this->emitLabel(CondLabel); if (!this->visitBool(Cond)) return false; @@ -4279,13 +4304,16 @@ bool Compiler<Emitter>::visitCXXForRangeStmt(const CXXForRangeStmt *S) { if (!this->visitLoopBody(Body)) return false; + this->fallthrough(IncLabel); this->emitLabel(IncLabel); if (!this->discard(Inc)) return false; } + if (!this->jump(CondLabel)) return false; + this->fallthrough(EndLabel); this->emitLabel(EndLabel); return true; } @@ -4991,7 +5019,7 @@ bool Compiler<Emitter>::visitDeclRef(const ValueDecl *D, const Expr *E) { if ((VD->hasGlobalStorage() || VD->isLocalVarDecl() || VD->isStaticDataMember()) && typeShouldBeVisited(VD->getType())) { - auto VarState = this->visitVarDecl(VD); + auto VarState = this->visitVarDecl(VD, true); if (VarState.notCreated()) return true; if (!VarState) @@ -5004,7 +5032,7 @@ bool Compiler<Emitter>::visitDeclRef(const ValueDecl *D, const Expr *E) { if (const auto *VD = dyn_cast<VarDecl>(D); VD && VD->getAnyInitializer() && VD->getType().isConstant(Ctx.getASTContext()) && !VD->isWeak()) { - auto VarState = this->visitVarDecl(VD); + auto VarState = this->visitVarDecl(VD, true); if (VarState.notCreated()) return true; if (!VarState) diff --git a/clang/lib/AST/Interp/Compiler.h b/clang/lib/AST/Interp/Compiler.h index d188ce2..d28526c 100644 --- a/clang/lib/AST/Interp/Compiler.h +++ b/clang/lib/AST/Interp/Compiler.h @@ -39,6 +39,7 @@ template <class Emitter> class SourceLocScope; template <class Emitter> class LoopScope; template <class Emitter> class LabelScope; template <class Emitter> class SwitchScope; +template <class Emitter> class StmtExprScope; template <class Emitter> class Compiler; struct InitLink { @@ -260,7 +261,7 @@ protected: /// intact. bool delegate(const Expr *E); /// Creates and initializes a variable from the given decl. - VarCreationState visitVarDecl(const VarDecl *VD); + VarCreationState visitVarDecl(const VarDecl *VD, bool Toplevel = false); /// Visit an APValue. bool visitAPValue(const APValue &Val, PrimType ValType, const Expr *E); bool visitAPValueInitializer(const APValue &Val, const Expr *E); @@ -334,6 +335,7 @@ private: friend class LoopScope<Emitter>; friend class LabelScope<Emitter>; friend class SwitchScope<Emitter>; + friend class StmtExprScope<Emitter>; /// Emits a zero initializer. bool visitZeroInitializer(PrimType T, QualType QT, const Expr *E); @@ -398,6 +400,8 @@ protected: /// Flag indicating if return value is to be discarded. bool DiscardResult = false; + bool InStmtExpr = false; + /// Flag inidicating if we're initializing an already created /// variable. This is set in visitInitializer(). bool Initializing = false; diff --git a/clang/lib/AST/Interp/Interp.h b/clang/lib/AST/Interp/Interp.h index ff6d50a..328db21 100644 --- a/clang/lib/AST/Interp/Interp.h +++ b/clang/lib/AST/Interp/Interp.h @@ -2618,6 +2618,13 @@ inline bool Invalid(InterpState &S, CodePtr OpPC) { return false; } +inline bool Unsupported(InterpState &S, CodePtr OpPC) { + const SourceLocation &Loc = S.Current->getLocation(OpPC); + S.FFDiag(Loc, diag::note_constexpr_stmt_expr_unsupported) + << S.Current->getRange(OpPC); + return false; +} + /// Do nothing and just abort execution. inline bool Error(InterpState &S, CodePtr OpPC) { return false; } diff --git a/clang/lib/AST/Interp/Opcodes.td b/clang/lib/AST/Interp/Opcodes.td index 81e7b81..8d01fe1 100644 --- a/clang/lib/AST/Interp/Opcodes.td +++ b/clang/lib/AST/Interp/Opcodes.td @@ -714,6 +714,7 @@ def Dup : Opcode { // [] -> [] def Invalid : Opcode {} +def Unsupported : Opcode {} def Error : Opcode {} def InvalidCast : Opcode { let Args = [ArgCastKind]; diff --git a/clang/lib/Basic/LangOptions.cpp b/clang/lib/Basic/LangOptions.cpp index 2b90646..61072b7 100644 --- a/clang/lib/Basic/LangOptions.cpp +++ b/clang/lib/Basic/LangOptions.cpp @@ -112,6 +112,7 @@ void LangOptions::setLangDefaults(LangOptions &Opts, Language Lang, Opts.C11 = Std.isC11(); Opts.C17 = Std.isC17(); Opts.C23 = Std.isC23(); + Opts.C2y = Std.isC2y(); Opts.CPlusPlus = Std.isCPlusPlus(); Opts.CPlusPlus11 = Std.isCPlusPlus11(); Opts.CPlusPlus14 = Std.isCPlusPlus14(); diff --git a/clang/lib/Basic/OpenMPKinds.cpp b/clang/lib/Basic/OpenMPKinds.cpp index df172dc..766d6a8 100644 --- a/clang/lib/Basic/OpenMPKinds.cpp +++ b/clang/lib/Basic/OpenMPKinds.cpp @@ -747,105 +747,85 @@ void clang::getOpenMPCaptureRegions( assert(unsigned(DKind) < llvm::omp::Directive_enumSize); assert(isOpenMPCapturingDirective(DKind) && "Expecting capturing directive"); - switch (DKind) { - case OMPD_metadirective: - CaptureRegions.push_back(OMPD_metadirective); - break; - case OMPD_parallel: - case OMPD_parallel_for: - case OMPD_parallel_for_simd: - case OMPD_parallel_master: - case OMPD_parallel_masked: - case OMPD_parallel_sections: - case OMPD_distribute_parallel_for: - case OMPD_distribute_parallel_for_simd: - case OMPD_parallel_loop: - CaptureRegions.push_back(OMPD_parallel); - break; - case OMPD_target_teams: - case OMPD_target_teams_distribute: - case OMPD_target_teams_distribute_simd: - CaptureRegions.push_back(OMPD_task); - CaptureRegions.push_back(OMPD_target); - CaptureRegions.push_back(OMPD_teams); - break; - case OMPD_teams: - case OMPD_teams_distribute: - case OMPD_teams_distribute_simd: - CaptureRegions.push_back(OMPD_teams); - break; - case OMPD_target: - case OMPD_target_simd: - CaptureRegions.push_back(OMPD_task); - CaptureRegions.push_back(OMPD_target); - break; - case OMPD_teams_loop: - case OMPD_teams_distribute_parallel_for: - case OMPD_teams_distribute_parallel_for_simd: - CaptureRegions.push_back(OMPD_teams); - CaptureRegions.push_back(OMPD_parallel); - break; - case OMPD_target_parallel: - case OMPD_target_parallel_for: - case OMPD_target_parallel_for_simd: - case OMPD_target_parallel_loop: - CaptureRegions.push_back(OMPD_task); - CaptureRegions.push_back(OMPD_target); - CaptureRegions.push_back(OMPD_parallel); - break; - case OMPD_task: - case OMPD_target_enter_data: - case OMPD_target_exit_data: - case OMPD_target_update: - CaptureRegions.push_back(OMPD_task); - break; - case OMPD_taskloop: - case OMPD_taskloop_simd: - case OMPD_master_taskloop: - case OMPD_master_taskloop_simd: - case OMPD_masked_taskloop: - case OMPD_masked_taskloop_simd: - CaptureRegions.push_back(OMPD_taskloop); - break; - case OMPD_parallel_masked_taskloop: - case OMPD_parallel_masked_taskloop_simd: - case OMPD_parallel_master_taskloop: - case OMPD_parallel_master_taskloop_simd: - CaptureRegions.push_back(OMPD_parallel); - CaptureRegions.push_back(OMPD_taskloop); - break; - case OMPD_target_teams_loop: - case OMPD_target_teams_distribute_parallel_for: - case OMPD_target_teams_distribute_parallel_for_simd: - CaptureRegions.push_back(OMPD_task); - CaptureRegions.push_back(OMPD_target); - CaptureRegions.push_back(OMPD_teams); - CaptureRegions.push_back(OMPD_parallel); - break; - case OMPD_nothing: - CaptureRegions.push_back(OMPD_nothing); - break; - case OMPD_loop: - // TODO: 'loop' may require different capture regions depending on the bind - // clause or the parent directive when there is no bind clause. Use - // OMPD_unknown for now. - case OMPD_simd: - case OMPD_for: - case OMPD_for_simd: - case OMPD_sections: - case OMPD_single: - case OMPD_taskgroup: - case OMPD_distribute: - case OMPD_ordered: - case OMPD_target_data: - case OMPD_distribute_simd: - case OMPD_scope: - case OMPD_dispatch: + auto GetRegionsForLeaf = [&](OpenMPDirectiveKind LKind) { + assert(isLeafConstruct(LKind) && "Epecting leaf directive"); + // Whether a leaf would require OMPD_unknown if it occured on its own. + switch (LKind) { + case OMPD_metadirective: + CaptureRegions.push_back(OMPD_metadirective); + break; + case OMPD_nothing: + CaptureRegions.push_back(OMPD_nothing); + break; + case OMPD_parallel: + CaptureRegions.push_back(OMPD_parallel); + break; + case OMPD_target: + CaptureRegions.push_back(OMPD_task); + CaptureRegions.push_back(OMPD_target); + break; + case OMPD_task: + case OMPD_target_enter_data: + case OMPD_target_exit_data: + case OMPD_target_update: + CaptureRegions.push_back(OMPD_task); + break; + case OMPD_teams: + CaptureRegions.push_back(OMPD_teams); + break; + case OMPD_taskloop: + CaptureRegions.push_back(OMPD_taskloop); + break; + case OMPD_loop: + // TODO: 'loop' may require different capture regions depending on the + // bind clause or the parent directive when there is no bind clause. + // If any of the directives that push regions here are parents of 'loop', + // assume 'parallel'. Otherwise do nothing. + if (!CaptureRegions.empty() && + !llvm::is_contained(CaptureRegions, OMPD_parallel)) + CaptureRegions.push_back(OMPD_parallel); + else + return true; + break; + case OMPD_dispatch: + case OMPD_distribute: + case OMPD_for: + case OMPD_masked: + case OMPD_master: + case OMPD_ordered: + case OMPD_scope: + case OMPD_sections: + case OMPD_simd: + case OMPD_single: + case OMPD_target_data: + case OMPD_taskgroup: + // These directives (when standalone) use OMPD_unknown as the region, + // but when they're constituents of a compound directive, and other + // leafs from that directive have specific regions, then these directives + // add no additional regions. + return true; + default: + llvm::errs() << getOpenMPDirectiveName(LKind) << '\n'; + llvm_unreachable("Unexpected directive"); + } + return false; + }; + + bool MayNeedUnknownRegion = false; + for (OpenMPDirectiveKind L : getLeafConstructsOrSelf(DKind)) + MayNeedUnknownRegion |= GetRegionsForLeaf(L); + + // We need OMPD_unknown when no regions were added, and specific leaf + // constructs were present. Push a single OMPD_unknown as the capture + /// region. + if (CaptureRegions.empty() && MayNeedUnknownRegion) CaptureRegions.push_back(OMPD_unknown); - break; - default: - llvm_unreachable("Unhandled OpenMP directive"); - } + + // OMPD_unknown is only expected as the only region. If other regions + // are present OMPD_unknown should not be present. + assert((CaptureRegions[0] == OMPD_unknown || + !llvm::is_contained(CaptureRegions, OMPD_unknown)) && + "Misplaced OMPD_unknown"); } bool clang::checkFailClauseParameter(OpenMPClauseKind FailClauseParameter) { diff --git a/clang/lib/Basic/Targets/NVPTX.h b/clang/lib/Basic/Targets/NVPTX.h index a5daf36..9a985e4 100644 --- a/clang/lib/Basic/Targets/NVPTX.h +++ b/clang/lib/Basic/Targets/NVPTX.h @@ -105,6 +105,7 @@ public: case 'l': case 'f': case 'd': + case 'q': Info.setAllowsRegister(); return true; } diff --git a/clang/lib/CodeGen/CGDeclCXX.cpp b/clang/lib/CodeGen/CGDeclCXX.cpp index e18b339..05dd7dd 100644 --- a/clang/lib/CodeGen/CGDeclCXX.cpp +++ b/clang/lib/CodeGen/CGDeclCXX.cpp @@ -840,6 +840,10 @@ CodeGenModule::EmitCXXGlobalInitFunc() { // No Itanium initializer in header like modules. if (M->isHeaderLikeModule()) continue; + // We're allowed to skip the initialization if we are sure it doesn't + // do any thing. + if (!M->isNamedModuleInterfaceHasInit()) + continue; llvm::FunctionType *FTy = llvm::FunctionType::get(VoidTy, false); SmallString<256> FnName; { diff --git a/clang/lib/CodeGen/CGVTables.cpp b/clang/lib/CodeGen/CGVTables.cpp index 3e88cd7..10d972e 100644 --- a/clang/lib/CodeGen/CGVTables.cpp +++ b/clang/lib/CodeGen/CGVTables.cpp @@ -1079,33 +1079,38 @@ CodeGenModule::getVTableLinkage(const CXXRecordDecl *RD) { if (!RD->isExternallyVisible()) return llvm::GlobalVariable::InternalLinkage; - // V-tables for non-template classes with an owning module are always - // uniquely emitted in that module. - if (RD->isInNamedModule()) - return llvm::GlobalVariable::ExternalLinkage; - - // We're at the end of the translation unit, so the current key - // function is fully correct. - const CXXMethodDecl *keyFunction = Context.getCurrentKeyFunction(RD); - if (keyFunction && !RD->hasAttr<DLLImportAttr>()) { + bool IsInNamedModule = RD->isInNamedModule(); + // If the CXXRecordDecl are not in a module unit, we need to get + // its key function. We're at the end of the translation unit, so the current + // key function is fully correct. + const CXXMethodDecl *keyFunction = + IsInNamedModule ? nullptr : Context.getCurrentKeyFunction(RD); + if (IsInNamedModule || (keyFunction && !RD->hasAttr<DLLImportAttr>())) { // If this class has a key function, use that to determine the // linkage of the vtable. const FunctionDecl *def = nullptr; - if (keyFunction->hasBody(def)) + if (keyFunction && keyFunction->hasBody(def)) keyFunction = cast<CXXMethodDecl>(def); - switch (keyFunction->getTemplateSpecializationKind()) { - case TSK_Undeclared: - case TSK_ExplicitSpecialization: + bool IsExternalDefinition = + IsInNamedModule ? RD->shouldEmitInExternalSource() : !def; + + TemplateSpecializationKind Kind = + IsInNamedModule ? RD->getTemplateSpecializationKind() + : keyFunction->getTemplateSpecializationKind(); + + switch (Kind) { + case TSK_Undeclared: + case TSK_ExplicitSpecialization: assert( - (def || CodeGenOpts.OptimizationLevel > 0 || + (IsInNamedModule || def || CodeGenOpts.OptimizationLevel > 0 || CodeGenOpts.getDebugInfo() != llvm::codegenoptions::NoDebugInfo) && - "Shouldn't query vtable linkage without key function, " - "optimizations, or debug info"); - if (!def && CodeGenOpts.OptimizationLevel > 0) + "Shouldn't query vtable linkage without the class in module units, " + "key function, optimizations, or debug info"); + if (IsExternalDefinition && CodeGenOpts.OptimizationLevel > 0) return llvm::GlobalVariable::AvailableExternallyLinkage; - if (keyFunction->isInlined()) + if (keyFunction && keyFunction->isInlined()) return !Context.getLangOpts().AppleKext ? llvm::GlobalVariable::LinkOnceODRLinkage : llvm::Function::InternalLinkage; @@ -1124,7 +1129,7 @@ CodeGenModule::getVTableLinkage(const CXXRecordDecl *RD) { case TSK_ExplicitInstantiationDeclaration: llvm_unreachable("Should not have been asked to emit this"); - } + } } // -fapple-kext mode does not support weak linkage, so we must use diff --git a/clang/lib/Driver/ToolChain.cpp b/clang/lib/Driver/ToolChain.cpp index 8f4cc47..977e083 100644 --- a/clang/lib/Driver/ToolChain.cpp +++ b/clang/lib/Driver/ToolChain.cpp @@ -811,6 +811,12 @@ std::optional<std::string> ToolChain::getStdlibPath() const { return getTargetSubDirPath(P); } +std::optional<std::string> ToolChain::getStdlibIncludePath() const { + SmallString<128> P(D.Dir); + llvm::sys::path::append(P, "..", "include"); + return getTargetSubDirPath(P); +} + ToolChain::path_list ToolChain::getArchSpecificLibPaths() const { path_list Paths; diff --git a/clang/lib/Driver/ToolChains/BareMetal.cpp b/clang/lib/Driver/ToolChains/BareMetal.cpp index dd365e6..11f9487 100644 --- a/clang/lib/Driver/ToolChains/BareMetal.cpp +++ b/clang/lib/Driver/ToolChains/BareMetal.cpp @@ -270,15 +270,19 @@ void BareMetal::AddClangSystemIncludeArgs(const ArgList &DriverArgs, addSystemInclude(DriverArgs, CC1Args, Dir.str()); } - if (!DriverArgs.hasArg(options::OPT_nostdlibinc)) { - const SmallString<128> SysRoot(computeSysRoot()); - if (!SysRoot.empty()) { - for (const Multilib &M : getOrderedMultilibs()) { - SmallString<128> Dir(SysRoot); - llvm::sys::path::append(Dir, M.includeSuffix()); - llvm::sys::path::append(Dir, "include"); - addSystemInclude(DriverArgs, CC1Args, Dir.str()); - } + if (DriverArgs.hasArg(options::OPT_nostdlibinc)) + return; + + if (std::optional<std::string> Path = getStdlibIncludePath()) + addSystemInclude(DriverArgs, CC1Args, *Path); + + const SmallString<128> SysRoot(computeSysRoot()); + if (!SysRoot.empty()) { + for (const Multilib &M : getOrderedMultilibs()) { + SmallString<128> Dir(SysRoot); + llvm::sys::path::append(Dir, M.includeSuffix()); + llvm::sys::path::append(Dir, "include"); + addSystemInclude(DriverArgs, CC1Args, Dir.str()); } } } @@ -296,6 +300,40 @@ void BareMetal::AddClangCXXStdlibIncludeArgs(const ArgList &DriverArgs, return; const Driver &D = getDriver(); + std::string Target = getTripleString(); + + auto AddCXXIncludePath = [&](StringRef Path) { + std::string Version = detectLibcxxVersion(Path); + if (Version.empty()) + return; + + { + // First the per-target include dir: include/<target>/c++/v1. + SmallString<128> TargetDir(Path); + llvm::sys::path::append(TargetDir, Target, "c++", Version); + addSystemInclude(DriverArgs, CC1Args, TargetDir); + } + + { + // Then the generic dir: include/c++/v1. + SmallString<128> Dir(Path); + llvm::sys::path::append(Dir, "c++", Version); + addSystemInclude(DriverArgs, CC1Args, Dir); + } + }; + + switch (GetCXXStdlibType(DriverArgs)) { + case ToolChain::CST_Libcxx: { + SmallString<128> P(D.Dir); + llvm::sys::path::append(P, "..", "include"); + AddCXXIncludePath(P); + break; + } + case ToolChain::CST_Libstdcxx: + // We only support libc++ toolchain installation. + break; + } + std::string SysRoot(computeSysRoot()); if (SysRoot.empty()) return; diff --git a/clang/lib/Driver/ToolChains/Linux.cpp b/clang/lib/Driver/ToolChains/Linux.cpp index 49e029e..98a878e 100644 --- a/clang/lib/Driver/ToolChains/Linux.cpp +++ b/clang/lib/Driver/ToolChains/Linux.cpp @@ -568,16 +568,12 @@ std::string Linux::getDynamicLinker(const ArgList &Args) const { Loader = (tools::ppc::hasPPCAbiArg(Args, "elfv1")) ? "ld64.so.1" : "ld64.so.2"; break; - case llvm::Triple::riscv32: { - StringRef ABIName = tools::riscv::getRISCVABI(Args, Triple); - LibDir = "lib"; - Loader = ("ld-linux-riscv32-" + ABIName + ".so.1").str(); - break; - } + case llvm::Triple::riscv32: case llvm::Triple::riscv64: { + StringRef ArchName = llvm::Triple::getArchTypeName(Arch); StringRef ABIName = tools::riscv::getRISCVABI(Args, Triple); LibDir = "lib"; - Loader = ("ld-linux-riscv64-" + ABIName + ".so.1").str(); + Loader = ("ld-linux-" + ArchName + "-" + ABIName + ".so.1").str(); break; } case llvm::Triple::sparc: diff --git a/clang/lib/Format/Format.cpp b/clang/lib/Format/Format.cpp index 259ef1d..6a8883b 100644 --- a/clang/lib/Format/Format.cpp +++ b/clang/lib/Format/Format.cpp @@ -729,6 +729,7 @@ template <> struct MappingTraits<FormatStyle::SpacesInLineComment> { template <> struct MappingTraits<FormatStyle::SpacesInParensCustom> { static void mapping(IO &IO, FormatStyle::SpacesInParensCustom &Spaces) { + IO.mapOptional("ExceptDoubleParentheses", Spaces.ExceptDoubleParentheses); IO.mapOptional("InCStyleCasts", Spaces.InCStyleCasts); IO.mapOptional("InConditionalStatements", Spaces.InConditionalStatements); IO.mapOptional("InEmptyParentheses", Spaces.InEmptyParentheses); @@ -1184,8 +1185,8 @@ template <> struct MappingTraits<FormatStyle> { (SpacesInParentheses || SpaceInEmptyParentheses || SpacesInConditionalStatement || SpacesInCStyleCastParentheses)) { if (SpacesInParentheses) { - // set all options except InCStyleCasts and InEmptyParentheses - // to true for backward compatibility. + // For backward compatibility. + Style.SpacesInParensOptions.ExceptDoubleParentheses = false; Style.SpacesInParensOptions.InConditionalStatements = true; Style.SpacesInParensOptions.InCStyleCasts = SpacesInCStyleCastParentheses; diff --git a/clang/lib/Format/TokenAnnotator.cpp b/clang/lib/Format/TokenAnnotator.cpp index 03082cd..0fd0214 100644 --- a/clang/lib/Format/TokenAnnotator.cpp +++ b/clang/lib/Format/TokenAnnotator.cpp @@ -4361,6 +4361,15 @@ bool TokenAnnotator::spaceRequiredBetween(const AnnotatedLine &Line, Right.is(tok::r_brace) && Right.isNot(BK_Block))) { return Style.SpacesInParensOptions.InEmptyParentheses; } + if (Style.SpacesInParens == FormatStyle::SIPO_Custom && + Style.SpacesInParensOptions.ExceptDoubleParentheses && + Left.is(tok::r_paren) && Right.is(tok::r_paren)) { + auto *InnerLParen = Left.MatchingParen; + if (InnerLParen && InnerLParen->Previous == Right.MatchingParen) { + InnerLParen->SpacesRequiredBefore = 0; + return false; + } + } if (Style.SpacesInParensOptions.InConditionalStatements) { const FormatToken *LeftParen = nullptr; if (Left.is(tok::l_paren)) diff --git a/clang/lib/Frontend/InitPreprocessor.cpp b/clang/lib/Frontend/InitPreprocessor.cpp index 55ec460..5e52555 100644 --- a/clang/lib/Frontend/InitPreprocessor.cpp +++ b/clang/lib/Frontend/InitPreprocessor.cpp @@ -448,7 +448,9 @@ static void InitializeStandardPredefinedMacros(const TargetInfo &TI, // value is, are implementation-defined. // (Removed in C++20.) if (!LangOpts.CPlusPlus) { - if (LangOpts.C23) + if (LangOpts.C2y) + Builder.defineMacro("__STDC_VERSION__", "202400L"); + else if (LangOpts.C23) Builder.defineMacro("__STDC_VERSION__", "202311L"); else if (LangOpts.C17) Builder.defineMacro("__STDC_VERSION__", "201710L"); diff --git a/clang/lib/Frontend/Rewrite/RewriteModernObjC.cpp b/clang/lib/Frontend/Rewrite/RewriteModernObjC.cpp index 6ae955a..764cdc6 100644 --- a/clang/lib/Frontend/Rewrite/RewriteModernObjC.cpp +++ b/clang/lib/Frontend/Rewrite/RewriteModernObjC.cpp @@ -465,15 +465,15 @@ namespace { std::string SynthesizeByrefCopyDestroyHelper(VarDecl *VD, int flag); std::string SynthesizeBlockHelperFuncs(BlockExpr *CE, int i, - StringRef funcName, std::string Tag); - std::string SynthesizeBlockFunc(BlockExpr *CE, int i, - StringRef funcName, std::string Tag); - std::string SynthesizeBlockImpl(BlockExpr *CE, - std::string Tag, std::string Desc); - std::string SynthesizeBlockDescriptor(std::string DescTag, - std::string ImplTag, - int i, StringRef funcName, - unsigned hasCopy); + StringRef funcName, + const std::string &Tag); + std::string SynthesizeBlockFunc(BlockExpr *CE, int i, StringRef funcName, + const std::string &Tag); + std::string SynthesizeBlockImpl(BlockExpr *CE, const std::string &Tag, + const std::string &Desc); + std::string SynthesizeBlockDescriptor(const std::string &DescTag, + const std::string &ImplTag, int i, + StringRef funcName, unsigned hasCopy); Stmt *SynthesizeBlockCall(CallExpr *Exp, const Expr* BlockExp); void SynthesizeBlockLiterals(SourceLocation FunLocStart, StringRef FunName); @@ -4037,7 +4037,7 @@ static bool HasLocalVariableExternalStorage(ValueDecl *VD) { std::string RewriteModernObjC::SynthesizeBlockFunc(BlockExpr *CE, int i, StringRef funcName, - std::string Tag) { + const std::string &Tag) { const FunctionType *AFT = CE->getFunctionType(); QualType RT = AFT->getReturnType(); std::string StructRef = "struct " + Tag; @@ -4131,9 +4131,8 @@ std::string RewriteModernObjC::SynthesizeBlockFunc(BlockExpr *CE, int i, return S; } -std::string RewriteModernObjC::SynthesizeBlockHelperFuncs(BlockExpr *CE, int i, - StringRef funcName, - std::string Tag) { +std::string RewriteModernObjC::SynthesizeBlockHelperFuncs( + BlockExpr *CE, int i, StringRef funcName, const std::string &Tag) { std::string StructRef = "struct " + Tag; std::string S = "static void __"; @@ -4175,8 +4174,9 @@ std::string RewriteModernObjC::SynthesizeBlockHelperFuncs(BlockExpr *CE, int i, return S; } -std::string RewriteModernObjC::SynthesizeBlockImpl(BlockExpr *CE, std::string Tag, - std::string Desc) { +std::string RewriteModernObjC::SynthesizeBlockImpl(BlockExpr *CE, + const std::string &Tag, + const std::string &Desc) { std::string S = "\nstruct " + Tag; std::string Constructor = " " + Tag; @@ -4290,10 +4290,9 @@ std::string RewriteModernObjC::SynthesizeBlockImpl(BlockExpr *CE, std::string Ta return S; } -std::string RewriteModernObjC::SynthesizeBlockDescriptor(std::string DescTag, - std::string ImplTag, int i, - StringRef FunName, - unsigned hasCopy) { +std::string RewriteModernObjC::SynthesizeBlockDescriptor( + const std::string &DescTag, const std::string &ImplTag, int i, + StringRef FunName, unsigned hasCopy) { std::string S = "\nstatic struct " + DescTag; S += " {\n size_t reserved;\n"; diff --git a/clang/lib/Interpreter/CMakeLists.txt b/clang/lib/Interpreter/CMakeLists.txt index 9065f99..6a06965 100644 --- a/clang/lib/Interpreter/CMakeLists.txt +++ b/clang/lib/Interpreter/CMakeLists.txt @@ -12,6 +12,10 @@ set(LLVM_LINK_COMPONENTS TargetParser ) +if (EMSCRIPTEN AND "lld" IN_LIST LLVM_ENABLE_PROJECTS) + set(WASM_SRC Wasm.cpp) +endif() + add_clang_library(clangInterpreter DeviceOffload.cpp CodeCompletion.cpp @@ -20,6 +24,8 @@ add_clang_library(clangInterpreter Interpreter.cpp InterpreterUtils.cpp Value.cpp + ${WASM_SRC} + PARTIAL_SOURCES_INTENDED DEPENDS intrinsics_gen diff --git a/clang/lib/Interpreter/IncrementalExecutor.cpp b/clang/lib/Interpreter/IncrementalExecutor.cpp index 6f03610..1824a5b 100644 --- a/clang/lib/Interpreter/IncrementalExecutor.cpp +++ b/clang/lib/Interpreter/IncrementalExecutor.cpp @@ -36,6 +36,8 @@ LLVM_ATTRIBUTE_USED void linkComponents() { } namespace clang { +IncrementalExecutor::IncrementalExecutor(llvm::orc::ThreadSafeContext &TSC) + : TSCtx(TSC) {} llvm::Expected<std::unique_ptr<llvm::orc::LLJITBuilder>> IncrementalExecutor::createDefaultJITBuilder( diff --git a/clang/lib/Interpreter/IncrementalExecutor.h b/clang/lib/Interpreter/IncrementalExecutor.h index b434720..7954cde 100644 --- a/clang/lib/Interpreter/IncrementalExecutor.h +++ b/clang/lib/Interpreter/IncrementalExecutor.h @@ -43,16 +43,19 @@ class IncrementalExecutor { llvm::DenseMap<const PartialTranslationUnit *, llvm::orc::ResourceTrackerSP> ResourceTrackers; +protected: + IncrementalExecutor(llvm::orc::ThreadSafeContext &TSC); + public: enum SymbolNameKind { IRName, LinkerName }; IncrementalExecutor(llvm::orc::ThreadSafeContext &TSC, llvm::orc::LLJITBuilder &JITBuilder, llvm::Error &Err); - ~IncrementalExecutor(); + virtual ~IncrementalExecutor(); - llvm::Error addModule(PartialTranslationUnit &PTU); - llvm::Error removeModule(PartialTranslationUnit &PTU); - llvm::Error runCtors() const; + virtual llvm::Error addModule(PartialTranslationUnit &PTU); + virtual llvm::Error removeModule(PartialTranslationUnit &PTU); + virtual llvm::Error runCtors() const; llvm::Error cleanUp(); llvm::Expected<llvm::orc::ExecutorAddr> getSymbolAddress(llvm::StringRef Name, SymbolNameKind NameKind) const; diff --git a/clang/lib/Interpreter/Interpreter.cpp b/clang/lib/Interpreter/Interpreter.cpp index 7a95278..49dc92d 100644 --- a/clang/lib/Interpreter/Interpreter.cpp +++ b/clang/lib/Interpreter/Interpreter.cpp @@ -15,6 +15,9 @@ #include "IncrementalExecutor.h" #include "IncrementalParser.h" #include "InterpreterUtils.h" +#ifdef __EMSCRIPTEN__ +#include "Wasm.h" +#endif // __EMSCRIPTEN__ #include "clang/AST/ASTContext.h" #include "clang/AST/Mangle.h" @@ -186,6 +189,12 @@ IncrementalCompilerBuilder::CreateCpp() { std::vector<const char *> Argv; Argv.reserve(5 + 1 + UserArgs.size()); Argv.push_back("-xc++"); +#ifdef __EMSCRIPTEN__ + Argv.push_back("-target"); + Argv.push_back("wasm32-unknown-emscripten"); + Argv.push_back("-pie"); + Argv.push_back("-shared"); +#endif Argv.insert(Argv.end(), UserArgs.begin(), UserArgs.end()); std::string TT = TargetTriple ? *TargetTriple : llvm::sys::getProcessTriple(); @@ -426,8 +435,12 @@ llvm::Error Interpreter::CreateExecutor() { } llvm::Error Err = llvm::Error::success(); +#ifdef __EMSCRIPTEN__ + auto Executor = std::make_unique<WasmIncrementalExecutor>(*TSCtx); +#else auto Executor = std::make_unique<IncrementalExecutor>(*TSCtx, *JITBuilder, Err); +#endif if (!Err) IncrExecutor = std::move(Executor); diff --git a/clang/lib/Interpreter/Wasm.cpp b/clang/lib/Interpreter/Wasm.cpp new file mode 100644 index 0000000..1001410 --- /dev/null +++ b/clang/lib/Interpreter/Wasm.cpp @@ -0,0 +1,114 @@ +//===----------------- Wasm.cpp - Wasm Interpreter --------------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// +// +// This file implements interpreter support for code execution in WebAssembly. +// +//===----------------------------------------------------------------------===// + +#include "Wasm.h" +#include "IncrementalExecutor.h" + +#include <llvm/IR/LegacyPassManager.h> +#include <llvm/IR/Module.h> +#include <llvm/MC/TargetRegistry.h> +#include <llvm/Target/TargetMachine.h> + +#include <clang/Interpreter/Interpreter.h> + +#include <string> + +namespace lld { +namespace wasm { +bool link(llvm::ArrayRef<const char *> args, llvm::raw_ostream &stdoutOS, + llvm::raw_ostream &stderrOS, bool exitEarly, bool disableOutput); +} // namespace wasm +} // namespace lld + +#include <dlfcn.h> + +namespace clang { + +WasmIncrementalExecutor::WasmIncrementalExecutor( + llvm::orc::ThreadSafeContext &TSC) + : IncrementalExecutor(TSC) {} + +llvm::Error WasmIncrementalExecutor::addModule(PartialTranslationUnit &PTU) { + std::string ErrorString; + + const llvm::Target *Target = llvm::TargetRegistry::lookupTarget( + PTU.TheModule->getTargetTriple(), ErrorString); + if (!Target) { + return llvm::make_error<llvm::StringError>("Failed to create Wasm Target: ", + llvm::inconvertibleErrorCode()); + } + + llvm::TargetOptions TO = llvm::TargetOptions(); + llvm::TargetMachine *TargetMachine = Target->createTargetMachine( + PTU.TheModule->getTargetTriple(), "", "", TO, llvm::Reloc::Model::PIC_); + PTU.TheModule->setDataLayout(TargetMachine->createDataLayout()); + std::string OutputFileName = PTU.TheModule->getName().str() + ".wasm"; + + std::error_code Error; + llvm::raw_fd_ostream OutputFile(llvm::StringRef(OutputFileName), Error); + + llvm::legacy::PassManager PM; + if (TargetMachine->addPassesToEmitFile(PM, OutputFile, nullptr, + llvm::CodeGenFileType::ObjectFile)) { + return llvm::make_error<llvm::StringError>( + "Wasm backend cannot produce object.", llvm::inconvertibleErrorCode()); + } + + if (!PM.run(*PTU.TheModule)) { + + return llvm::make_error<llvm::StringError>("Failed to emit Wasm object.", + llvm::inconvertibleErrorCode()); + } + + OutputFile.close(); + + std::vector<const char *> LinkerArgs = {"wasm-ld", + "-pie", + "--import-memory", + "--no-entry", + "--export-all", + "--experimental-pic", + "--no-export-dynamic", + "--stack-first", + OutputFileName.c_str(), + "-o", + OutputFileName.c_str()}; + int Result = + lld::wasm::link(LinkerArgs, llvm::outs(), llvm::errs(), false, false); + if (!Result) + return llvm::make_error<llvm::StringError>( + "Failed to link incremental module", llvm::inconvertibleErrorCode()); + + void *LoadedLibModule = + dlopen(OutputFileName.c_str(), RTLD_NOW | RTLD_GLOBAL); + if (LoadedLibModule == nullptr) { + llvm::errs() << dlerror() << '\n'; + return llvm::make_error<llvm::StringError>( + "Failed to load incremental module", llvm::inconvertibleErrorCode()); + } + + return llvm::Error::success(); +} + +llvm::Error WasmIncrementalExecutor::removeModule(PartialTranslationUnit &PTU) { + return llvm::make_error<llvm::StringError>("Not implemented yet", + llvm::inconvertibleErrorCode()); +} + +llvm::Error WasmIncrementalExecutor::runCtors() const { + // This seems to be automatically done when using dlopen() + return llvm::Error::success(); +} + +WasmIncrementalExecutor::~WasmIncrementalExecutor() = default; + +} // namespace clang diff --git a/clang/lib/Interpreter/Wasm.h b/clang/lib/Interpreter/Wasm.h new file mode 100644 index 0000000..b1fd880 --- /dev/null +++ b/clang/lib/Interpreter/Wasm.h @@ -0,0 +1,37 @@ +//===------------------ Wasm.h - Wasm Interpreter ---------------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// +// +// This file implements interpreter support for code execution in WebAssembly. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_CLANG_LIB_INTERPRETER_WASM_H +#define LLVM_CLANG_LIB_INTERPRETER_WASM_H + +#ifndef __EMSCRIPTEN__ +#error "This requires emscripten." +#endif // __EMSCRIPTEN__ + +#include "IncrementalExecutor.h" + +namespace clang { + +class WasmIncrementalExecutor : public IncrementalExecutor { +public: + WasmIncrementalExecutor(llvm::orc::ThreadSafeContext &TSC); + + llvm::Error addModule(PartialTranslationUnit &PTU) override; + llvm::Error removeModule(PartialTranslationUnit &PTU) override; + llvm::Error runCtors() const override; + + ~WasmIncrementalExecutor() override; +}; + +} // namespace clang + +#endif // LLVM_CLANG_LIB_INTERPRETER_WASM_H diff --git a/clang/lib/Parse/ParseExpr.cpp b/clang/lib/Parse/ParseExpr.cpp index 9fc3cd7..2dd2e1b 100644 --- a/clang/lib/Parse/ParseExpr.cpp +++ b/clang/lib/Parse/ParseExpr.cpp @@ -3449,7 +3449,8 @@ ExprResult Parser::ParseGenericSelectionExpression() { } const auto *LIT = cast<LocInfoType>(ControllingType.get().get()); SourceLocation Loc = LIT->getTypeSourceInfo()->getTypeLoc().getBeginLoc(); - Diag(Loc, diag::ext_generic_with_type_arg); + Diag(Loc, getLangOpts().C2y ? diag::warn_c2y_compat_generic_with_type_arg + : diag::ext_c2y_generic_with_type_arg); } else { // C11 6.5.1.1p3 "The controlling expression of a generic selection is // not evaluated." diff --git a/clang/lib/Sema/CheckExprLifetime.cpp b/clang/lib/Sema/CheckExprLifetime.cpp index be77949..2f9ef28 100644 --- a/clang/lib/Sema/CheckExprLifetime.cpp +++ b/clang/lib/Sema/CheckExprLifetime.cpp @@ -966,27 +966,11 @@ static bool pathOnlyInitializesGslPointer(IndirectLocalPath &Path) { return false; } -static void checkExprLifetimeImpl( - Sema &SemaRef, - llvm::PointerUnion<const InitializedEntity *, const AssignedEntity *> - CEntity, - Expr *Init) { - LifetimeKind LK = LK_FullExpression; - - const AssignedEntity *AEntity = nullptr; - // Local variables for initialized entity. - const InitializedEntity *InitEntity = nullptr; - const InitializedEntity *ExtendingEntity = nullptr; - if (CEntity.is<const InitializedEntity *>()) { - InitEntity = CEntity.get<const InitializedEntity *>(); - auto LTResult = getEntityLifetime(InitEntity); - LK = LTResult.getInt(); - ExtendingEntity = LTResult.getPointer(); - } else { - AEntity = CEntity.get<const AssignedEntity *>(); - if (AEntity->LHS->getType()->isPointerType()) // builtin pointer type - LK = LK_Extended; - } +static void checkExprLifetimeImpl(Sema &SemaRef, + const InitializedEntity *InitEntity, + const InitializedEntity *ExtendingEntity, + LifetimeKind LK, + const AssignedEntity *AEntity, Expr *Init) { // If this entity doesn't have an interesting lifetime, don't bother looking // for temporaries within its initializer. if (LK == LK_FullExpression) @@ -1024,7 +1008,19 @@ static void checkExprLifetimeImpl( return true; } } + if (AEntity) { + if (!MTE) + return false; + assert(shouldLifetimeExtendThroughPath(Path) == + PathLifetimeKind::NoExtend && + "No lifetime extension for assignments"); + if (!pathContainsInit(Path)) + SemaRef.Diag(DiagLoc, diag::warn_dangling_pointer_assignment) + << AEntity->LHS << DiagRange; + return false; + } + assert(InitEntity && "only for initialization"); switch (LK) { case LK_FullExpression: llvm_unreachable("already handled this"); @@ -1047,8 +1043,6 @@ static void checkExprLifetimeImpl( switch (shouldLifetimeExtendThroughPath(Path)) { case PathLifetimeKind::Extend: - assert(InitEntity && "Lifetime extension should happen only for " - "initialization and not assignment"); // Update the storage duration of the materialized temporary. // FIXME: Rebuild the expression instead of mutating it. MTE->setExtendingDecl(ExtendingEntity->getDecl(), @@ -1057,8 +1051,6 @@ static void checkExprLifetimeImpl( return true; case PathLifetimeKind::ShouldExtend: - assert(InitEntity && "Lifetime extension should happen only for " - "initialization and not assignment"); // We're supposed to lifetime-extend the temporary along this path (per // the resolution of DR1815), but we don't support that yet. // @@ -1076,23 +1068,16 @@ static void checkExprLifetimeImpl( if (pathContainsInit(Path)) return false; - if (InitEntity) { - SemaRef.Diag(DiagLoc, diag::warn_dangling_variable) - << RK << !InitEntity->getParent() - << ExtendingEntity->getDecl()->isImplicit() - << ExtendingEntity->getDecl() << Init->isGLValue() << DiagRange; - } else { - SemaRef.Diag(DiagLoc, diag::warn_dangling_pointer_assignment) - << AEntity->LHS << DiagRange; - } + SemaRef.Diag(DiagLoc, diag::warn_dangling_variable) + << RK << !InitEntity->getParent() + << ExtendingEntity->getDecl()->isImplicit() + << ExtendingEntity->getDecl() << Init->isGLValue() << DiagRange; break; } break; } case LK_MemInitializer: { - assert(InitEntity && "Expect only on initializing the entity"); - if (MTE) { // Under C++ DR1696, if a mem-initializer (or a default member // initializer used by the absence of one) would lifetime-extend a @@ -1167,7 +1152,6 @@ static void checkExprLifetimeImpl( } case LK_New: - assert(InitEntity && "Expect only on initializing the entity"); if (isa<MaterializeTemporaryExpr>(L)) { if (IsGslPtrInitWithGslTempOwner) SemaRef.Diag(DiagLoc, diag::warn_dangling_lifetime_pointer) @@ -1185,7 +1169,6 @@ static void checkExprLifetimeImpl( case LK_Return: case LK_StmtExprResult: - assert(InitEntity && "Expect only on initializing the entity"); if (auto *DRE = dyn_cast<DeclRefExpr>(L)) { // We can't determine if the local variable outlives the statement // expression. @@ -1292,12 +1275,18 @@ static void checkExprLifetimeImpl( void checkExprLifetime(Sema &SemaRef, const InitializedEntity &Entity, Expr *Init) { - checkExprLifetimeImpl(SemaRef, &Entity, Init); + auto LTResult = getEntityLifetime(&Entity); + LifetimeKind LK = LTResult.getInt(); + const InitializedEntity *ExtendingEntity = LTResult.getPointer(); + checkExprLifetimeImpl(SemaRef, &Entity, ExtendingEntity, LK, nullptr, Init); } void checkExprLifetime(Sema &SemaRef, const AssignedEntity &Entity, Expr *Init) { - checkExprLifetimeImpl(SemaRef, &Entity, Init); + LifetimeKind LK = LK_FullExpression; + if (Entity.LHS->getType()->isPointerType()) // builtin pointer type + LK = LK_Extended; + checkExprLifetimeImpl(SemaRef, nullptr, nullptr, LK, &Entity, Init); } } // namespace clang::sema diff --git a/clang/lib/Sema/SemaExpr.cpp b/clang/lib/Sema/SemaExpr.cpp index cc811a4..852344d 100644 --- a/clang/lib/Sema/SemaExpr.cpp +++ b/clang/lib/Sema/SemaExpr.cpp @@ -4486,8 +4486,16 @@ bool Sema::CheckUnaryExprOrTypeTraitOperand(QualType ExprType, // When alignof or _Alignof is applied to an array type, the result // is the alignment of the element type. if (ExprKind == UETT_AlignOf || ExprKind == UETT_PreferredAlignOf || - ExprKind == UETT_OpenMPRequiredSimdAlign) + ExprKind == UETT_OpenMPRequiredSimdAlign) { + // If the trait is 'alignof' in C before C2y, the ability to apply the + // trait to an incomplete array is an extension. + if (ExprKind == UETT_AlignOf && !getLangOpts().CPlusPlus && + ExprType->isIncompleteArrayType()) + Diag(OpLoc, getLangOpts().C2y + ? diag::warn_c2y_compat_alignof_incomplete_array + : diag::ext_c2y_alignof_incomplete_array); ExprType = Context.getBaseElementType(ExprType); + } if (ExprKind == UETT_VecStep) return CheckVecStepTraitOperandType(*this, ExprType, OpLoc, ExprRange); @@ -13764,8 +13772,9 @@ static QualType CheckIncrementDecrementOperand(Sema &S, Expr *Op, return QualType(); } else if (ResType->isAnyComplexType()) { // C99 does not support ++/-- on complex types, we allow as an extension. - S.Diag(OpLoc, diag::ext_increment_complex) - << IsInc << Op->getSourceRange(); + S.Diag(OpLoc, S.getLangOpts().C2y ? diag::warn_c2y_compat_increment_complex + : diag::ext_c2y_increment_complex) + << IsInc << Op->getSourceRange(); } else if (ResType->isPlaceholderType()) { ExprResult PR = S.CheckPlaceholderExpr(Op); if (PR.isInvalid()) return QualType(); diff --git a/clang/lib/Sema/SemaOverload.cpp b/clang/lib/Sema/SemaOverload.cpp index 982cca6..5ea6b06 100644 --- a/clang/lib/Sema/SemaOverload.cpp +++ b/clang/lib/Sema/SemaOverload.cpp @@ -40,6 +40,7 @@ #include "llvm/ADT/DenseSet.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/STLForwardCompat.h" +#include "llvm/ADT/ScopeExit.h" #include "llvm/ADT/SmallPtrSet.h" #include "llvm/ADT/SmallString.h" #include "llvm/ADT/SmallVector.h" @@ -11884,6 +11885,46 @@ static void DiagnoseFailedExplicitSpec(Sema &S, OverloadCandidate *Cand) { << (ES.getExpr() ? ES.getExpr()->getSourceRange() : SourceRange()); } +static void NoteImplicitDeductionGuide(Sema &S, FunctionDecl *Fn) { + auto *DG = dyn_cast<CXXDeductionGuideDecl>(Fn); + if (!DG) + return; + TemplateDecl *OriginTemplate = + DG->getDeclName().getCXXDeductionGuideTemplate(); + // We want to always print synthesized deduction guides for type aliases. + // They would retain the explicit bit of the corresponding constructor. + if (!(DG->isImplicit() || (OriginTemplate && OriginTemplate->isTypeAlias()))) + return; + std::string FunctionProto; + llvm::raw_string_ostream OS(FunctionProto); + FunctionTemplateDecl *Template = DG->getDescribedFunctionTemplate(); + if (!Template) { + // This also could be an instantiation. Find out the primary template. + FunctionDecl *Pattern = + DG->getTemplateInstantiationPattern(/*ForDefinition=*/false); + if (!Pattern) { + // The implicit deduction guide is built on an explicit non-template + // deduction guide. Currently, this might be the case only for type + // aliases. + // FIXME: Add a test once https://github.com/llvm/llvm-project/pull/96686 + // gets merged. + assert(OriginTemplate->isTypeAlias() && + "Non-template implicit deduction guides are only possible for " + "type aliases"); + DG->print(OS); + S.Diag(DG->getLocation(), diag::note_implicit_deduction_guide) + << FunctionProto; + return; + } + Template = Pattern->getDescribedFunctionTemplate(); + assert(Template && "Cannot find the associated function template of " + "CXXDeductionGuideDecl?"); + } + Template->print(OS); + S.Diag(DG->getLocation(), diag::note_implicit_deduction_guide) + << FunctionProto; +} + /// Generates a 'note' diagnostic for an overload candidate. We've /// already generated a primary error at the call site. /// @@ -11941,6 +11982,17 @@ static void NoteFunctionCandidate(Sema &S, OverloadCandidate *Cand, return; } + // If this is a synthesized deduction guide we're deducing against, add a note + // for it. These deduction guides are not explicitly spelled in the source + // code, so simply printing a deduction failure note mentioning synthesized + // template parameters or pointing to the header of the surrounding RecordDecl + // would be confusing. + // + // We prefer adding such notes at the end of the deduction failure because + // duplicate code snippets appearing in the diagnostic would likely become + // noisy. + auto _ = llvm::make_scope_exit([&] { NoteImplicitDeductionGuide(S, Fn); }); + switch (Cand->FailureKind) { case ovl_fail_too_many_arguments: case ovl_fail_too_few_arguments: diff --git a/clang/test/AST/Interp/c.c b/clang/test/AST/Interp/c.c index 684658d..9ec305d 100644 --- a/clang/test/AST/Interp/c.c +++ b/clang/test/AST/Interp/c.c @@ -293,3 +293,11 @@ void SuperSpecialFunc(void) { const int SuperSpecialCase = 10; _Static_assert((sizeof(SuperSpecialCase) == 12 && SuperSpecialCase == 3) || SuperSpecialCase == 10, ""); // pedantic-warning {{GNU extension}} } + + +void T1(void) { + static int *y[1] = {({ static int _x = 20; (void*)0;})}; // all-error {{initializer element is not a compile-time constant}} \ + // pedantic-warning {{use of GNU statement expression extension}} +} + + diff --git a/clang/test/AST/Interp/literals.cpp b/clang/test/AST/Interp/literals.cpp index a7a602e..f70ca79e 100644 --- a/clang/test/AST/Interp/literals.cpp +++ b/clang/test/AST/Interp/literals.cpp @@ -1243,3 +1243,26 @@ namespace Extern { static_assert(&ExternNonLiteralVarDecl() == &nl, ""); #endif } + +#if __cplusplus >= 201402L +constexpr int StmtExprEval() { + if (({ + while (0); + true; + })) { + return 2; + } + return 1; +} +static_assert(StmtExprEval() == 2, ""); + +constexpr int ReturnInStmtExpr() { // both-error {{never produces a constant expression}} + return ({ + return 1; // both-note 2{{this use of statement expressions is not supported in a constant expression}} + 2; + }); +} +static_assert(ReturnInStmtExpr() == 1, ""); // both-error {{not an integral constant expression}} \ + // both-note {{in call to}} + +#endif diff --git a/clang/test/C/C2y/n3192.c b/clang/test/C/C2y/n3192.c new file mode 100644 index 0000000..b6f61d5 --- /dev/null +++ b/clang/test/C/C2y/n3192.c @@ -0,0 +1,43 @@ +// RUN: %clang_cc1 -verify -Wno-c23-extensions %s + +/* WG14 N3192: Yes + * Sequential hexdigits + */ + +// expected-no-diagnostics + +// Demonstrate that hex digits are already sequential in all targets Clang +// supports. + +#define TEST_VAL(ch) ((ch >= 'A' && ch <= 'F') || (ch >= 'a' && ch <= 'f')) +#define GET_VAL(ch) (((ch >= 'A' && ch <= 'F') ? (ch - 'A') : (ch - 'a')) + 10) + +_Static_assert(TEST_VAL('A')); +_Static_assert(TEST_VAL('B')); +_Static_assert(TEST_VAL('C')); +_Static_assert(TEST_VAL('D')); +_Static_assert(TEST_VAL('E')); +_Static_assert(TEST_VAL('F')); +_Static_assert(TEST_VAL('a')); +_Static_assert(TEST_VAL('b')); +_Static_assert(TEST_VAL('c')); +_Static_assert(TEST_VAL('d')); +_Static_assert(TEST_VAL('e')); +_Static_assert(TEST_VAL('f')); + +_Static_assert(!TEST_VAL('G')); +_Static_assert(!TEST_VAL('h')); + +_Static_assert(GET_VAL('A') == 0xA); +_Static_assert(GET_VAL('B') == 0xB); +_Static_assert(GET_VAL('C') == 0xC); +_Static_assert(GET_VAL('D') == 0xD); +_Static_assert(GET_VAL('E') == 0xE); +_Static_assert(GET_VAL('F') == 0xF); +_Static_assert(GET_VAL('a') == 0xA); +_Static_assert(GET_VAL('b') == 0xB); +_Static_assert(GET_VAL('c') == 0xC); +_Static_assert(GET_VAL('d') == 0xD); +_Static_assert(GET_VAL('e') == 0xE); +_Static_assert(GET_VAL('f') == 0xF); + diff --git a/clang/test/C/C2y/n3259.c b/clang/test/C/C2y/n3259.c new file mode 100644 index 0000000..42c6720 --- /dev/null +++ b/clang/test/C/C2y/n3259.c @@ -0,0 +1,66 @@ +// RUN: %clang_cc1 -triple=x86_64 -std=c2y -Wall -pedantic -Wno-unused -Wpre-c2y-compat -verify=pre-c2y %s -emit-llvm -o - | FileCheck %s +// RUN: %clang_cc1 -triple=x86_64 -std=c23 -Wall -pedantic -Wno-unused %s -verify -emit-llvm -o - | FileCheck %s + +/* WG14 N3259: Yes + * Support ++ and -- on complex values + */ + +// CHECK-LABEL: define {{.*}} void @test() +void test() { + // CHECK: %[[F:.+]] = alloca { float, float } + // CHECK: store float 1 + // CHECK: store float 0 + _Complex float f = __builtin_complex(1.0f, 0.0f); + + // CHECK: %[[F_REALP1:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_REAL:.+]] = load float, ptr %[[F_REALP1]] + // CHECK-NEXT: %[[F_IMAGP2:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: %[[F_IMAG:.+]] = load float, ptr %[[F_IMAGP2]] + // CHECK-NEXT: %[[INC:.+]] = fadd float %[[F_REAL]], 1.000000e+00 + // CHECK-NEXT: %[[F_REALP3:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_IMAGP4:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: store float %[[INC]], ptr %[[F_REALP3]] + // CHECK-NEXT: store float %[[F_IMAG]], ptr %[[F_IMAGP4]] + f++; /* expected-warning {{'++' on an object of complex type is a C2y extension}} + pre-c2y-warning {{'++' on an object of complex type is incompatible with C standards before C2y}} + */ + + // CHECK: %[[F_REALP5:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_REAL6:.+]] = load float, ptr %[[F_REALP5]] + // CHECK-NEXT: %[[F_IMAGP7:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: %[[F_IMAG8:.+]] = load float, ptr %[[F_IMAGP7]] + // CHECK-NEXT: %[[INC9:.+]] = fadd float %[[F_REAL6]], 1.000000e+00 + // CHECK-NEXT: %[[F_REALP10:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_IMAGP11:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: store float %[[INC9]], ptr %[[F_REALP10]] + // CHECK-NEXT: store float %[[F_IMAG8]], ptr %[[F_IMAGP11]] + ++f; /* expected-warning {{'++' on an object of complex type is a C2y extension}} + pre-c2y-warning {{'++' on an object of complex type is incompatible with C standards before C2y}} + */ + + // CHECK: %[[F_REALP12:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_REAL13:.+]] = load float, ptr %[[F_REALP12]] + // CHECK-NEXT: %[[F_IMAGP14:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: %[[F_IMAG15:.+]] = load float, ptr %[[F_IMAGP14]] + // CHECK-NEXT: %[[DEC:.+]] = fadd float %[[F_REAL13]], -1.000000e+00 + // CHECK-NEXT: %[[F_REALP16:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_IMAGP17:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: store float %[[DEC]], ptr %[[F_REALP16]] + // CHECK-NEXT: store float %[[F_IMAG15]], ptr %[[F_IMAGP17]] + f--; /* expected-warning {{'--' on an object of complex type is a C2y extension}} + pre-c2y-warning {{'--' on an object of complex type is incompatible with C standards before C2y}} + */ + + // CHECK: %[[F_REALP18:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_REAL19:.+]] = load float, ptr %[[F_REALP18]] + // CHECK-NEXT: %[[F_IMAGP20:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: %[[F_IMAG21:.+]] = load float, ptr %[[F_IMAGP20]] + // CHECK-NEXT: %[[DEC22:.+]] = fadd float %[[F_REAL19]], -1.000000e+00 + // CHECK-NEXT: %[[F_REALP23:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 0 + // CHECK-NEXT: %[[F_IMAGP24:.+]] = getelementptr inbounds { float, float }, ptr %[[F]], i32 0, i32 1 + // CHECK-NEXT: store float %[[DEC22]], ptr %[[F_REALP23]] + // CHECK-NEXT: store float %[[F_IMAG21]], ptr %[[F_IMAGP24]] + --f; /* expected-warning {{'--' on an object of complex type is a C2y extension}} + pre-c2y-warning {{'--' on an object of complex type is incompatible with C standards before C2y}} + */ +} diff --git a/clang/test/C/C2y/n3260.c b/clang/test/C/C2y/n3260.c new file mode 100644 index 0000000..e7b0eef --- /dev/null +++ b/clang/test/C/C2y/n3260.c @@ -0,0 +1,16 @@ +// RUN: %clang_cc1 -verify -std=c2y -Wall -pedantic -Wpre-c2y-compat %s +// RUN: %clang_cc1 -verify=pre-c2y -std=c23 -Wall -pedantic %s + +/* WG14 N3260: Clang 17 + * Generic selection expression with a type operand + */ + +static_assert( + _Generic( + const int, /* pre-c2y-warning {{passing a type argument as the first operand to '_Generic' is a C2y extension}} + expected-warning {{passing a type argument as the first operand to '_Generic' is incompatible with C standards before C2y}} + */ + int : 0, + const int : 1 + ) +); diff --git a/clang/test/C/C2y/n3273.c b/clang/test/C/C2y/n3273.c new file mode 100644 index 0000000..84eeadd --- /dev/null +++ b/clang/test/C/C2y/n3273.c @@ -0,0 +1,14 @@ +// RUN: %clang_cc1 -verify -std=c2y -Wall -pedantic -Wpre-c2y-compat %s +// RUN: %clang_cc1 -verify=pre-c2y -std=c23 -Wall -pedantic %s + +/* WG14 N3273: Clang 3.5 + * alignof of an incomplete array type + */ + +static_assert( + alignof(int[]) == /* pre-c2y-warning {{'alignof' on an incomplete array type is a C2y extension}} + expected-warning {{'alignof' on an incomplete array type is incompatible with C standards before C2y}} + */ + alignof(int) +); + diff --git a/clang/test/C/C99/n809.c b/clang/test/C/C99/n809.c index 7297443..48a72cc 100644 --- a/clang/test/C/C99/n809.c +++ b/clang/test/C/C99/n809.c @@ -108,11 +108,11 @@ void func(void) { // a real type; Clang supports this as an extension on complex types as well. _Complex float cf = 0.0f; - cf++; // expected-warning {{'++' on an object of complex type is a Clang extension}} - ++cf; // expected-warning {{'++' on an object of complex type is a Clang extension}} + cf++; // expected-warning {{'++' on an object of complex type is a C2y extension}} + ++cf; // expected-warning {{'++' on an object of complex type is a C2y extension}} - cf--; // expected-warning {{'--' on an object of complex type is a Clang extension}} - --cf; // expected-warning {{'--' on an object of complex type is a Clang extension}} + cf--; // expected-warning {{'--' on an object of complex type is a C2y extension}} + --cf; // expected-warning {{'--' on an object of complex type is a C2y extension}} // However, unary + and - are fine, as is += 1. (void)-cf; diff --git a/clang/test/CXX/drs/cwg26xx.cpp b/clang/test/CXX/drs/cwg26xx.cpp index e180c93..d843b09 100644 --- a/clang/test/CXX/drs/cwg26xx.cpp +++ b/clang/test/CXX/drs/cwg26xx.cpp @@ -193,8 +193,11 @@ static_assert(__is_same(decltype(i), I<char, 4>)); J j = { "ghi" }; // since-cxx20-error@-1 {{no viable constructor or deduction guide}} // since-cxx20-note@#cwg2681-J {{candidate template ignored: could not match 'J<N>' against 'const char *'}} +// since-cxx20-note@#cwg2681-J {{implicit deduction guide declared as 'template <size_t N> J(J<N>) -> J<N>'}} // since-cxx20-note@#cwg2681-J {{candidate template ignored: could not match 'const unsigned char' against 'const char'}} +// since-cxx20-note@#cwg2681-J {{implicit deduction guide declared as 'template <size_t N> J(const unsigned char (&)[N]) -> J<N>'}} // since-cxx20-note@#cwg2681-J {{candidate function template not viable: requires 0 arguments, but 1 was provided}} +// since-cxx20-note@#cwg2681-J {{implicit deduction guide declared as 'template <size_t N> J() -> J<N>'}} #endif } diff --git a/clang/test/CXX/expr/expr.post/expr.type.conv/p1.cpp b/clang/test/CXX/expr/expr.post/expr.type.conv/p1.cpp index f3608bc3..f8d88e1 100644 --- a/clang/test/CXX/expr/expr.post/expr.type.conv/p1.cpp +++ b/clang/test/CXX/expr/expr.post/expr.type.conv/p1.cpp @@ -1,6 +1,6 @@ // RUN: %clang_cc1 -std=c++1z -verify %s -template<typename T> struct A { // expected-note 2{{candidate}} +template<typename T> struct A { // expected-note 2{{candidate}} expected-note 2{{implicit deduction guide}} T t, u; }; template<typename T> A(T, T) -> A<T>; // expected-note {{deduced conflicting types for parameter 'T'}} diff --git a/clang/test/CXX/over/over.match/over.match.funcs/over.match.class.deduct/p2.cpp b/clang/test/CXX/over/over.match/over.match.funcs/over.match.class.deduct/p2.cpp index 49fde29..d192070 100644 --- a/clang/test/CXX/over/over.match/over.match.funcs/over.match.class.deduct/p2.cpp +++ b/clang/test/CXX/over/over.match/over.match.funcs/over.match.class.deduct/p2.cpp @@ -35,8 +35,9 @@ namespace std { } namespace p0702r1 { - template<typename T> struct X { // expected-note {{candidate}} - X(std::initializer_list<T>); // expected-note {{candidate template ignored: could not match 'std::initializer_list<T>' against 'Z'}} + template<typename T> struct X { // expected-note {{candidate}} expected-note {{implicit deduction guide}} + X(std::initializer_list<T>); // expected-note {{candidate template ignored: could not match 'std::initializer_list<T>' against 'Z'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T> X(std::initializer_list<T>) -> X<T>'}} }; X xi = {0}; diff --git a/clang/test/CXX/temp/temp.fct.spec/temp.deduct/temp.deduct.call/p3-0x.cpp b/clang/test/CXX/temp/temp.fct.spec/temp.deduct/temp.deduct.call/p3-0x.cpp index 8592626..ed44536 100644 --- a/clang/test/CXX/temp/temp.fct.spec/temp.deduct/temp.deduct.call/p3-0x.cpp +++ b/clang/test/CXX/temp/temp.fct.spec/temp.deduct/temp.deduct.call/p3-0x.cpp @@ -6,8 +6,8 @@ #if __cplusplus > 201402L namespace ClassTemplateParamNotForwardingRef { // This is not a forwarding reference. - template<typename T> struct A { // expected-note {{candidate}} - A(T&&); // expected-note {{expects an rvalue}} + template<typename T> struct A { // expected-note {{candidate}} expected-note {{implicit deduction guide}} + A(T&&); // expected-note {{expects an rvalue}} expected-note {{implicit deduction guide}} }; int n; A a = n; // expected-error {{no viable constructor or deduction guide}} @@ -75,10 +75,12 @@ namespace std_example { int n3 = g(i); // expected-error{{no matching function for call to 'g'}} #if __cplusplus > 201402L - template<class T> struct A { // expected-note {{candidate}} + template<class T> struct A { // expected-note {{candidate}} expected-note {{implicit deduction guide}} template<class U> - A(T &&, U &&, int *); // expected-note {{[with T = int, U = int] not viable: expects an rvalue}} - A(T &&, int *); // expected-note {{requires 2}} + A(T &&, U &&, int *); // expected-note {{[with T = int, U = int] not viable: expects an rvalue}} \ + // expected-note {{implicit deduction guide declared as 'template <class T, class U> A(T &&, type-parameter-0-1 &&, int *) -> A<T>'}} + A(T &&, int *); // expected-note {{requires 2}} \ + // expected-note {{implicit deduction guide declared as 'template <class T> A(T &&, int *) -> A<T>'}} }; template<class T> A(T &&, int *) -> A<T>; // expected-note {{requires 2}} diff --git a/clang/test/Driver/Inputs/basic_baremetal_tree/bin/.keep b/clang/test/Driver/Inputs/basic_baremetal_tree/bin/.keep new file mode 100644 index 0000000..e69de29 --- /dev/null +++ b/clang/test/Driver/Inputs/basic_baremetal_tree/bin/.keep diff --git a/clang/test/Driver/Inputs/basic_baremetal_tree/include/armv6m-unknown-none-eabi/.keep b/clang/test/Driver/Inputs/basic_baremetal_tree/include/armv6m-unknown-none-eabi/.keep new file mode 100644 index 0000000..e69de29 --- /dev/null +++ b/clang/test/Driver/Inputs/basic_baremetal_tree/include/armv6m-unknown-none-eabi/.keep diff --git a/clang/test/Driver/Inputs/basic_baremetal_tree/include/armv6m-unknown-none-eabi/c++/v1/.keep b/clang/test/Driver/Inputs/basic_baremetal_tree/include/armv6m-unknown-none-eabi/c++/v1/.keep new file mode 100644 index 0000000..e69de29 --- /dev/null +++ b/clang/test/Driver/Inputs/basic_baremetal_tree/include/armv6m-unknown-none-eabi/c++/v1/.keep diff --git a/clang/test/Driver/Inputs/basic_baremetal_tree/include/c++/v1/.keep b/clang/test/Driver/Inputs/basic_baremetal_tree/include/c++/v1/.keep new file mode 100644 index 0000000..e69de29 --- /dev/null +++ b/clang/test/Driver/Inputs/basic_baremetal_tree/include/c++/v1/.keep diff --git a/clang/test/Driver/Inputs/basic_baremetal_tree/lib/armv6m-unknown-none-eabi/.keep b/clang/test/Driver/Inputs/basic_baremetal_tree/lib/armv6m-unknown-none-eabi/.keep new file mode 100644 index 0000000..e69de29 --- /dev/null +++ b/clang/test/Driver/Inputs/basic_baremetal_tree/lib/armv6m-unknown-none-eabi/.keep diff --git a/clang/test/Driver/baremetal.cpp b/clang/test/Driver/baremetal.cpp index cc14f04..de4e934 100644 --- a/clang/test/Driver/baremetal.cpp +++ b/clang/test/Driver/baremetal.cpp @@ -13,7 +13,7 @@ // CHECK-V6M-C-SAME: "-resource-dir" "[[RESOURCE_DIR:[^"]+]]" // CHECK-V6M-C-SAME: "-isysroot" "[[SYSROOT:[^"]*]]" // CHECK-V6M-C-SAME: "-internal-isystem" "[[SYSROOT]]{{[/\\]+}}include{{[/\\]+}}c++{{[/\\]+}}v1" -// CHECk-V6M-C-SAME: "-internal-isystem" "[[SYSROOT]]{{[/\\]+}}include" +// CHECK-V6M-C-SAME: "-internal-isystem" "[[SYSROOT]]{{[/\\]+}}include" // CHECK-V6M-C-SAME: "-x" "c++" "{{.*}}baremetal.cpp" // CHECK-V6M-C-NEXT: ld{{(.exe)?}}" "{{.*}}.o" "-Bstatic" "-EL" // CHECK-V6M-C-SAME: "-T" "semihosted.lds" "-Lsome{{[/\\]+}}directory{{[/\\]+}}user{{[/\\]+}}asked{{[/\\]+}}for" @@ -26,6 +26,20 @@ // RUN: --sysroot=%S/Inputs/baremetal_arm | FileCheck --check-prefix=CHECK-V6M-LIBINC %s // CHECK-V6M-LIBINC-NOT: "-internal-isystem" +// RUN: %clang %s -### --target=armv6m-none-eabi -o %t.out 2>&1 \ +// RUN: -ccc-install-dir %S/Inputs/basic_baremetal_tree/bin \ +// RUN: | FileCheck --check-prefix=CHECK-V6M-TREE %s +// CHECK-V6M-TREE: InstalledDir: [[INSTALLED_DIR:.+]] +// CHECK-V6M-TREE: "-cc1" "-triple" "thumbv6m-unknown-none-eabi" +// CHECK-V6M-TREE-SAME: "-resource-dir" "[[RESOURCE_DIR:[^"]+]]" +// CHECK-V6M-TREE-SAME: "-internal-isystem" "[[INSTALLED_DIR]]{{[/\\]+}}..{{[/\\]+}}include{{[/\\]+}}armv6m-unknown-none-eabi{{[/\\]+}}c++{{[/\\]+}}v1" +// CHECK-V6M-TREE-SAME: {{^}} "-internal-isystem" "[[INSTALLED_DIR]]{{[/\\]+}}..{{[/\\]+}}include{{[/\\]+}}c++{{[/\\]+}}v1" +// CHECK-V6M-TREE-SAME: "-internal-isystem" "[[INSTALLED_DIR]]{{[/\\]+}}..{{[/\\]+}}include{{[/\\]+}}armv6m-unknown-none-eabi" +// CHECK-V6M-TREE-SAME: "-x" "c++" "{{.*}}baremetal.cpp" +// CHECK-V6M-TREE-NEXT: ld{{(.exe)?}}" "{{.*}}.o" "-Bstatic" "-EL" +// CHECK-V6M-TREE-SAME: "-L[[INSTALLED_DIR]]{{[/\\]+}}..{{[/\\]+}}lib{{[/\\]+}}armv6m-unknown-none-eabi" +// CHECK-V6M-TREE-SAME: "-lc" "-lm" "{{[^"]*}}libclang_rt.builtins.a" "--target2=rel" "-o" "{{.*}}.tmp.out" + // RUN: %clang %s -### --target=armv7m-vendor-none-eabi -rtlib=compiler-rt 2>&1 \ // RUN: --sysroot=%S/Inputs/baremetal_arm \ // RUN: -resource-dir=%S/Inputs/resource_dir_with_per_target_subdir \ diff --git a/clang/test/Driver/unknown-std.c b/clang/test/Driver/unknown-std.c index 564c633..8f9047b 100644 --- a/clang/test/Driver/unknown-std.c +++ b/clang/test/Driver/unknown-std.c @@ -18,6 +18,8 @@ // CHECK-NEXT: note: use 'gnu17' or 'gnu18' for 'ISO C 2017 with GNU extensions' standard // CHECK-NEXT: note: use 'c23' for 'Working Draft for ISO C23' standard // CHECK-NEXT: note: use 'gnu23' for 'Working Draft for ISO C23 with GNU extensions' standard +// CHECK-NEXT: note: use 'c2y' for 'Working Draft for ISO C2y' standard +// CHECK-NEXT: note: use 'gnu2y' for 'Working Draft for ISO C2y with GNU extensions' standard // Make sure that no other output is present. // CHECK-NOT: {{^.+$}} diff --git a/clang/test/Misc/warning-flags.c b/clang/test/Misc/warning-flags.c index ae14720..cdbe1e9 100644 --- a/clang/test/Misc/warning-flags.c +++ b/clang/test/Misc/warning-flags.c @@ -88,4 +88,4 @@ CHECK-NEXT: warn_weak_import The list of warnings in -Wpedantic should NEVER grow. -CHECK: Number in -Wpedantic (not covered by other -W flags): 26 +CHECK: Number in -Wpedantic (not covered by other -W flags): 25 diff --git a/clang/test/Modules/pr97244.cppm b/clang/test/Modules/pr97244.cppm new file mode 100644 index 0000000..ad8b466 --- /dev/null +++ b/clang/test/Modules/pr97244.cppm @@ -0,0 +1,30 @@ +// REQUIRES: !system-windows +// +// RUN: rm -rf %t +// RUN: mkdir -p %t +// RUN: split-file %s %t +// +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Empty.cppm \ +// RUN: -emit-module-interface -o %t/Empty.pcm +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Empty2.cppm \ +// RUN: -fprebuilt-module-path=%t -emit-module-interface -o %t/Empty2.pcm +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/main.cpp \ +// RUN: -fprebuilt-module-path=%t -emit-llvm -o - | FileCheck %t/main.cpp +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Empty2.pcm \ +// RUN: -fprebuilt-module-path=%t -emit-llvm -o - | FileCheck %t/Empty2.cppm + +//--- Empty.cppm +export module Empty; + +//--- Empty2.cppm +export module Empty2; +import Empty; + +// CHECK-NOT: _ZGIW5Empty + +//--- main.cpp +import Empty; +import Empty2; + +// CHECK-NOT: _ZGIW5Empty +// CHECK-NOT: _ZGIW6Empty2 diff --git a/clang/test/Modules/pr97313.cppm b/clang/test/Modules/pr97313.cppm new file mode 100644 index 0000000..ebbd0ee --- /dev/null +++ b/clang/test/Modules/pr97313.cppm @@ -0,0 +1,118 @@ +// REQUIRES: !system-windows +// +// RUN: rm -rf %t +// RUN: mkdir -p %t +// RUN: split-file %s %t +// +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Base.cppm \ +// RUN: -emit-module-interface -o %t/Base.pcm +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Sub.cppm \ +// RUN: -emit-module-interface -o %t/Sub.pcm -fprebuilt-module-path=%t +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Sub.pcm \ +// RUN: -emit-llvm -o %t/Sub.pcm -o - -fprebuilt-module-path=%t | \ +// RUN: FileCheck %t/Sub.cppm +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/main.cpp \ +// RUN: -emit-llvm -fprebuilt-module-path=%t -o - | FileCheck %t/main.cpp +// +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Mod.cppm \ +// RUN: -emit-module-interface -o %t/Mod.pcm +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Mod.pcm \ +// RUN: -emit-llvm -o - | FileCheck %t/Mod.cppm +// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/Use.cpp \ +// RUN: -emit-llvm -fprebuilt-module-path=%t -o - | \ +// RUN: FileCheck %t/Use.cpp + +//--- Base.cppm +export module Base; + +export template <class> +class Base +{ +public: + constexpr Base(); + constexpr virtual ~Base(); +}; + +template <class X> +constexpr Base<X>::Base() = default; + +template <class X> +constexpr Base<X>::~Base() = default; + +//--- Sub.cppm +export module Sub; +export import Base; + +export class Sub : public Base<int> +{ +}; + +// CHECK: @_ZTIW4Base4BaseIiE = {{.*}}linkonce_odr + +//--- main.cpp +import Sub; + +int main() +{ + Base<int> *b = new Sub(); + delete b; +} + +// CHECK: @_ZTIW4Base4BaseIiE = {{.*}}linkonce_odr + +//--- Mod.cppm +export module Mod; + +export class NonTemplate { +public: + virtual ~NonTemplate(); +}; + +// CHECK: @_ZTIW3Mod11NonTemplate = {{.*}}constant + +export template <class C> +class Template { +public: + virtual ~Template(); +}; + +export template<> +class Template<char> { +public: + virtual ~Template(); +}; + +// CHECK: @_ZTIW3Mod8TemplateIcE = {{.*}}constant + +export template class Template<unsigned>; + +// CHECK: @_ZTIW3Mod8TemplateIjE = {{.*}}weak_odr + +export extern template class Template<double>; + +auto v = new Template<signed int>(); + +// CHECK: @_ZTIW3Mod8TemplateIiE = {{.*}}linkonce_odr + +//--- Use.cpp +import Mod; + +auto v1 = new NonTemplate(); +auto v2 = new Template<char>(); +auto v3 = new Template<unsigned>(); +auto v4 = new Template<double>(); +auto v5 = new Template<signed int>(); +auto v6 = new Template<NonTemplate>(); + +// CHECK: @_ZTVW3Mod11NonTemplate = {{.*}}external +// CHECK: @_ZTVW3Mod8TemplateIcE = {{.*}}external +// CHECK: @_ZTVW3Mod8TemplateIjE = {{.*}}weak_odr +// CHECK: @_ZTSW3Mod8TemplateIjE = {{.*}}weak_odr +// CHECK: @_ZTIW3Mod8TemplateIjE = {{.*}}weak_odr +// CHECK: @_ZTVW3Mod8TemplateIdE = {{.*}}external +// CHECK: @_ZTVW3Mod8TemplateIiE = {{.*}}linkonce_odr +// CHECK: @_ZTSW3Mod8TemplateIiE = {{.*}}linkonce_odr +// CHECK: @_ZTIW3Mod8TemplateIiE = {{.*}}linkonce_odr +// CHECK: @_ZTVW3Mod8TemplateIS_11NonTemplateE = {{.*}}linkonce_odr +// CHECK: @_ZTSW3Mod8TemplateIS_11NonTemplateE = {{.*}}linkonce_odr +// CHECK: @_ZTIW3Mod8TemplateIS_11NonTemplateE = {{.*}}linkonce_odr diff --git a/clang/test/Modules/template_name_lookup.cpp b/clang/test/Modules/template_name_lookup.cpp index 29375e5..82b06e8 100644 --- a/clang/test/Modules/template_name_lookup.cpp +++ b/clang/test/Modules/template_name_lookup.cpp @@ -7,5 +7,7 @@ import foo; void use() { X x; // expected-error {{no viable constructor or deduction guide for deduction of template arguments of 'X'}} // expected-note@Inputs/template_name_lookup/foo.cppm:3 {{candidate template ignored: couldn't infer template argument 'T'}} + // expected-note@Inputs/template_name_lookup/foo.cppm:3 {{implicit deduction guide declared as 'template <typename T> X(X<T>) -> X<T>'}} // expected-note@Inputs/template_name_lookup/foo.cppm:3 {{candidate function template not viable: requires 1 argument, but 0 were provided}} + // expected-note@Inputs/template_name_lookup/foo.cppm:3 {{implicit deduction guide declared as 'template <typename T> X() -> X<T>'}} } diff --git a/clang/test/PCH/cxx-explicit-specifier.cpp b/clang/test/PCH/cxx-explicit-specifier.cpp index 94ca9f7..84548fa 100644 --- a/clang/test/PCH/cxx-explicit-specifier.cpp +++ b/clang/test/PCH/cxx-explicit-specifier.cpp @@ -79,8 +79,8 @@ struct A { B<true> b_true; B<false> b_false; #else -//expected-note@-8 {{candidate template ignored}} -//expected-note@-8 {{explicit constructor declared here}} +//expected-note@-8 {{candidate template ignored}} expected-note@-8 {{implicit deduction guide declared as 'template <bool b> A(A<b>) -> A<b>'}} +//expected-note@-8 {{explicit constructor declared here}} expected-note@-8 {{implicit deduction guide declared as 'template <bool b> explicit(b) A(B<b>) -> A<b>'}} //expected-note@-15+ {{candidate constructor}} //expected-note@-8+ {{explicit conversion function is not a candidate (explicit specifier}} //expected-note@-11 {{explicit constructor is not a candidate (explicit specifier}} diff --git a/clang/test/Parser/generic-selection-type-extension-pedantic.c b/clang/test/Parser/generic-selection-type-extension-pedantic.c index d73e809..e611703 100644 --- a/clang/test/Parser/generic-selection-type-extension-pedantic.c +++ b/clang/test/Parser/generic-selection-type-extension-pedantic.c @@ -4,7 +4,7 @@ // in the right location. void test(void) { (void)_Generic( - int, // expected-warning {{passing a type argument as the first operand to '_Generic' is a Clang extension}} + int, // expected-warning {{passing a type argument as the first operand to '_Generic' is a C2y extension}} int : 0); (void)_Generic( 12, diff --git a/clang/test/Preprocessor/c2y.c b/clang/test/Preprocessor/c2y.c new file mode 100644 index 0000000..1d659d2 --- /dev/null +++ b/clang/test/Preprocessor/c2y.c @@ -0,0 +1,6 @@ +// RUN: %clang_cc1 -fsyntax-only -verify -std=c2y %s +// expected-no-diagnostics + +// FIXME: Set this to test the correct value once that value is set by WG14. +static_assert(__STDC_VERSION__ > 202311L, "Incorrect __STDC_VERSION__"); + diff --git a/clang/test/Sema/complex-inc-dec.c b/clang/test/Sema/complex-inc-dec.c index ebc23c3..4c780da 100644 --- a/clang/test/Sema/complex-inc-dec.c +++ b/clang/test/Sema/complex-inc-dec.c @@ -5,20 +5,20 @@ void func(void) { _Complex double cd; _Complex long double cld; - ++cf; // expected-warning {{'++' on an object of complex type is a Clang extension}} - ++cd; // expected-warning {{'++' on an object of complex type is a Clang extension}} - ++cld; // expected-warning {{'++' on an object of complex type is a Clang extension}} + ++cf; // expected-warning {{'++' on an object of complex type is a C2y extension}} + ++cd; // expected-warning {{'++' on an object of complex type is a C2y extension}} + ++cld; // expected-warning {{'++' on an object of complex type is a C2y extension}} - --cf; // expected-warning {{'--' on an object of complex type is a Clang extension}} - --cd; // expected-warning {{'--' on an object of complex type is a Clang extension}} - --cld; // expected-warning {{'--' on an object of complex type is a Clang extension}} + --cf; // expected-warning {{'--' on an object of complex type is a C2y extension}} + --cd; // expected-warning {{'--' on an object of complex type is a C2y extension}} + --cld; // expected-warning {{'--' on an object of complex type is a C2y extension}} - cf++; // expected-warning {{'++' on an object of complex type is a Clang extension}} - cd++; // expected-warning {{'++' on an object of complex type is a Clang extension}} - cld++; // expected-warning {{'++' on an object of complex type is a Clang extension}} + cf++; // expected-warning {{'++' on an object of complex type is a C2y extension}} + cd++; // expected-warning {{'++' on an object of complex type is a C2y extension}} + cld++; // expected-warning {{'++' on an object of complex type is a C2y extension}} - cf--; // expected-warning {{'--' on an object of complex type is a Clang extension}} - cd--; // expected-warning {{'--' on an object of complex type is a Clang extension}} - cld--; // expected-warning {{'--' on an object of complex type is a Clang extension}} + cf--; // expected-warning {{'--' on an object of complex type is a C2y extension}} + cd--; // expected-warning {{'--' on an object of complex type is a C2y extension}} + cld--; // expected-warning {{'--' on an object of complex type is a C2y extension}} } diff --git a/clang/test/Sema/tls_alignment.cpp b/clang/test/Sema/tls_alignment.cpp index c5c79aa..65b6b03 100644 --- a/clang/test/Sema/tls_alignment.cpp +++ b/clang/test/Sema/tls_alignment.cpp @@ -22,7 +22,9 @@ struct struct_with_aligned_field { template <typename> struct templated_struct {}; // expected-note@-1{{candidate template ignored: couldn't infer template argument ''}} -// expected-note@-2{{candidate function template not viable: requires 1 argument, but 0 were provided}} +// expected-note@-2{{implicit deduction guide declared as 'template <typename> templated_struct() -> templated_struct<type-parameter-0-0>'}} +// expected-note@-3{{candidate function template not viable: requires 1 argument, but 0 were provided}} +// expected-note@-4{{implicit deduction guide declared as 'template <typename> templated_struct(templated_struct<type-parameter-0-0>) -> templated_struct<type-parameter-0-0>'}} // A typedef of the aligned struct. typedef aligned_struct another_aligned_struct; diff --git a/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp b/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp index 90404f1..9ef5303 100644 --- a/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp +++ b/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp @@ -139,11 +139,13 @@ namespace look_into_current_instantiation { // templates, and members of the current instantiation A<float> &r = a; - template<typename T> struct B { // expected-note {{could not match 'B<T>' against 'int'}} + template<typename T> struct B { // expected-note {{could not match 'B<T>' against 'int'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T> B(B<T>) -> B<T>'}} struct X { typedef T type; }; - B(typename X::type); // expected-note {{couldn't infer template argument 'T'}} + B(typename X::type); // expected-note {{couldn't infer template argument 'T'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T> B(typename X::type) -> B<T>'}} }; B b = 0; // expected-error {{no viable}} @@ -564,8 +566,10 @@ namespace PR47175 { // Ensure we don't crash when CTAD fails. template <typename T1, typename T2> -struct Foo { // expected-note{{candidate function template not viable}} - Foo(T1, T2); // expected-note{{candidate function template not viable}} +struct Foo { // expected-note {{candidate function template not viable}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T1, typename T2> Foo(Foo<T1, T2>) -> Foo<T1, T2>'}} + Foo(T1, T2); // expected-note {{candidate function template not viable}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T1, typename T2> Foo(T1, T2) -> Foo<T1, T2>'}} }; template <typename... Args> diff --git a/clang/test/SemaCXX/cxx20-ctad-type-alias.cpp b/clang/test/SemaCXX/cxx20-ctad-type-alias.cpp index b71dfc6..3dafe82 100644 --- a/clang/test/SemaCXX/cxx20-ctad-type-alias.cpp +++ b/clang/test/SemaCXX/cxx20-ctad-type-alias.cpp @@ -110,6 +110,8 @@ struct Foo { template <typename X, int Y> using Bar = Foo<X, sizeof(X)>; // expected-note {{candidate template ignored: couldn't infer template argument 'X'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename X> requires __is_deducible(test9::Bar, Foo<type-parameter-0-0, sizeof(type-parameter-0-0)>) Bar(Foo<type-parameter-0-0, sizeof(type-parameter-0-0)>) -> Foo<type-parameter-0-0, sizeof(type-parameter-0-0)>'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename X> requires __is_deducible(test9::Bar, Foo<type-parameter-0-0, sizeof(type-parameter-0-0)>) Bar(const type-parameter-0-0 (&)[sizeof(type-parameter-0-0)]) -> Foo<type-parameter-0-0, sizeof(type-parameter-0-0)>'}} \ // expected-note {{candidate template ignored: constraints not satisfied [with X = int]}} \ // expected-note {{cannot deduce template arguments for 'Bar' from 'Foo<int, 4UL>'}} @@ -137,9 +139,12 @@ struct A {}; template<class T> struct Foo { T c; }; template<class X, class Y=A> using AFoo = Foo<Y>; // expected-note {{candidate template ignored: could not match 'Foo<type-parameter-0-0>' against 'int'}} \ + // expected-note {{implicit deduction guide declared as 'template <class Y = A> requires __is_deducible(test11::AFoo, Foo<type-parameter-0-0>) AFoo(Foo<type-parameter-0-0>) -> Foo<type-parameter-0-0>'}} \ // expected-note {{candidate template ignored: constraints not satisfied [with Y = int]}} \ // expected-note {{cannot deduce template arguments for 'AFoo' from 'Foo<int>'}} \ - // expected-note {{candidate function template not viable: requires 0 arguments, but 1 was provided}} + // expected-note {{implicit deduction guide declared as 'template <class Y = A> requires __is_deducible(test11::AFoo, Foo<type-parameter-0-0>) AFoo(type-parameter-0-0) -> Foo<type-parameter-0-0>'}} \ + // expected-note {{candidate function template not viable: requires 0 arguments, but 1 was provided}} \ + // expected-note {{implicit deduction guide declared as 'template <class Y = A> requires __is_deducible(test11::AFoo, Foo<type-parameter-0-0>) AFoo() -> Foo<type-parameter-0-0>'}} AFoo s = {1}; // expected-error {{no viable constructor or deduction guide for deduction of template arguments of 'AFoo'}} } // namespace test11 @@ -192,6 +197,8 @@ struct Foo { template <int K> using Bar = Foo<double, K>; // expected-note {{constraints not satisfied for class template 'Foo'}} // expected-note@-1 {{candidate template ignored: could not match}} +// expected-note@-2 {{implicit deduction guide declared as 'template <int K> requires __is_deducible(test14::Bar, Foo<double, K>) Bar(Foo<double, K>) -> Foo<double, K>'}} +// expected-note@-3 {{implicit deduction guide declared as 'template <int K> requires __is_deducible(test14::Bar, Foo<double, K>) Bar(const double (&)[K]) -> Foo<double, K>'}} double abc[3]; Bar s2 = {abc}; // expected-error {{no viable constructor or deduction guide for deduction }} } // namespace test14 @@ -204,7 +211,9 @@ template<typename> concept False = false; template<False W> using BFoo = AFoo<W>; // expected-note {{candidate template ignored: constraints not satisfied [with V = int]}} \ // expected-note {{cannot deduce template arguments for 'BFoo' from 'Foo<int *>'}} \ - // expected-note {{candidate template ignored: could not match 'Foo<type-parameter-0-0 *>' against 'int *'}} + // expected-note {{implicit deduction guide declared as 'template <class V> requires __is_deducible(AFoo, Foo<type-parameter-0-0 *>) && __is_deducible(test15::BFoo, Foo<type-parameter-0-0 *>) BFoo(type-parameter-0-0 *) -> Foo<type-parameter-0-0 *>}} \ + // expected-note {{candidate template ignored: could not match 'Foo<type-parameter-0-0 *>' against 'int *'}} \ + // expected-note {{template <class V> requires __is_deducible(AFoo, Foo<type-parameter-0-0 *>) && __is_deducible(test15::BFoo, Foo<type-parameter-0-0 *>) BFoo(Foo<type-parameter-0-0 *>) -> Foo<type-parameter-0-0 *>}} int i = 0; AFoo a1(&i); // OK, deduce Foo<int *> @@ -255,8 +264,11 @@ Foo(T) -> Foo<int>; template <typename U> using Bar = Foo<U>; // expected-note {{could not match 'Foo<type-parameter-0-0>' against 'int'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename U> requires __is_deducible(test18::Bar, Foo<type-parameter-0-0>) Bar(Foo<type-parameter-0-0>) -> Foo<type-parameter-0-0>'}} \ // expected-note {{candidate template ignored: constraints not satisfied}} \ - // expected-note {{candidate function template not viable}} + // expected-note {{implicit deduction guide declared as 'template <typename T> requires False<T> && __is_deducible(test18::Bar, Foo<int>) Bar(type-parameter-0-0) -> Foo<int>'}} \ + // expected-note {{candidate function template not viable}} \ + // expected-note {{implicit deduction guide declared as 'template <typename U> requires __is_deducible(test18::Bar, Foo<type-parameter-0-0>) Bar() -> Foo<type-parameter-0-0>'}} Bar s = {1}; // expected-error {{no viable constructor or deduction guide for deduction of template arguments}} } // namespace test18 @@ -284,7 +296,8 @@ class Foo {}; // Verify that template template type parameter TTP is referenced/used in the // template arguments of the RHS. template <template<typename> typename TTP> -using Bar = Foo<K<TTP>>; // expected-note {{candidate template ignored: could not match 'Foo<K<template-parameter-0-0>>' against 'int'}} +using Bar = Foo<K<TTP>>; // expected-note {{candidate template ignored: could not match 'Foo<K<template-parameter-0-0>>' against 'int'}} \ + // expected-note {{implicit deduction guide declared as 'template <template <typename> typename TTP> requires __is_deducible(test20::Bar, Foo<K<template-parameter-0-0>>) Bar(Foo<K<template-parameter-0-0>>) -> Foo<K<template-parameter-0-0>>'}} template <class T> class Container {}; diff --git a/clang/test/SemaCXX/cxx2a-explicit-bool.cpp b/clang/test/SemaCXX/cxx2a-explicit-bool.cpp index 9fdc059..03799c5 100644 --- a/clang/test/SemaCXX/cxx2a-explicit-bool.cpp +++ b/clang/test/SemaCXX/cxx2a-explicit-bool.cpp @@ -394,9 +394,11 @@ using type = T; template<typename T1, typename T2, bool b> struct A { // expected-note@-1+ {{candidate function}} + // expected-note@-2+ {{implicit deduction guide}} explicit(false) A(typename nondeduced<T1>::type, typename nondeduced<T2>::type, typename nondeduced<B<b>>::type) {} // expected-note@-1+ {{candidate template ignored}} + // expected-note@-2+ {{implicit deduction guide}} }; template<typename T1, typename T2, bool b> @@ -678,10 +680,12 @@ namespace deduction_guide2 { template<typename T1 = int, typename T2 = int> struct A { // expected-note@-1+ {{candidate template ignored}} + // expected-note@-2+ {{implicit deduction guide}} explicit(!is_same<T1, T2>::value) A(T1 = 0, T2 = 0) {} // expected-note@-1 {{explicit constructor declared here}} // expected-note@-2 2{{explicit constructor is not a candidate}} + // expected-note@-3 2{{implicit deduction guide declared}} }; A a0 = 0; diff --git a/clang/test/SemaCXX/gh65522.cpp b/clang/test/SemaCXX/gh65522.cpp index 2d6331b..8028124 100644 --- a/clang/test/SemaCXX/gh65522.cpp +++ b/clang/test/SemaCXX/gh65522.cpp @@ -3,9 +3,11 @@ class X {}; template<typename T> -class B3 { // expected-note {{candidate template ignored: could not match 'B3<T>' against 'int'}} +class B3 { // expected-note {{candidate template ignored: could not match 'B3<T>' against 'int'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T> B3(B3<T>) -> B3<T>'}} template<X x> B3(T); // expected-warning 2{{non-type template parameter of type 'X' is incompatible with C++ standards before C++20}} \ - // expected-note {{candidate template ignored: couldn't infer template argument 'x'}} + // expected-note {{candidate template ignored: couldn't infer template argument 'x'}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T, X x> B3(T) -> B3<T>'}} }; B3 b3 = 0; // expected-error {{no viable constructor or deduction guide for deduction of template arguments of 'B3'}} \ // expected-note {{while building implicit deduction guide first needed here}} diff --git a/clang/test/SemaCXX/invalid-deduction-guide-as-template-candidates.cpp b/clang/test/SemaCXX/invalid-deduction-guide-as-template-candidates.cpp index e9b362d..ad7441f 100644 --- a/clang/test/SemaCXX/invalid-deduction-guide-as-template-candidates.cpp +++ b/clang/test/SemaCXX/invalid-deduction-guide-as-template-candidates.cpp @@ -1,6 +1,8 @@ // RUN: %clang_cc1 -std=c++17 -fsyntax-only -verify %s template <class T> class Foo {}; // expected-note {{candidate template ignored: couldn't infer template argument 'T'}} \ - // expected-note {{candidate function template not viable: requires 1 argument, but 0 were provided}} + // expected-note {{implicit deduction guide declared as 'template <class T> Foo(Foo<T>) -> Foo<T>'}} \ + // expected-note {{candidate function template not viable: requires 1 argument, but 0 were provided}} \ + // expected-note {{implicit deduction guide declared as 'template <class T> Foo() -> Foo<T>'}} Foo(); // expected-error {{deduction guide declaration without trailing return type}} Foo vs; // expected-error {{no viable constructor or deduction guide for deduction of template arguments of 'Foo'}} diff --git a/clang/test/SemaCXX/lookup-template-name-extern-CXX.cpp b/clang/test/SemaCXX/lookup-template-name-extern-CXX.cpp index 93e7fb6..f851af7 100644 --- a/clang/test/SemaCXX/lookup-template-name-extern-CXX.cpp +++ b/clang/test/SemaCXX/lookup-template-name-extern-CXX.cpp @@ -3,8 +3,10 @@ // RUN: %clang_cc1 -std=c++20 %s -fsyntax-only -verify extern "C++" { template <class T> -class X {}; // expected-note {{candidate template ignored: couldn't infer template argument 'T'}} - // expected-note@-1 {{candidate function template not viable: requires 1 argument, but 0 were provided}} +class X {}; // expected-note {{candidate template ignored: couldn't infer template argument 'T'}} \ + // expected-note {{implicit deduction guide declared as 'template <class T> X(X<T>) -> X<T>'}} \ + // expected-note {{candidate function template not viable: requires 1 argument, but 0 were provided}} \ + // expected-note {{implicit deduction guide declared as 'template <class T> X() -> X<T>'}} } void foo() { diff --git a/clang/test/SemaCXX/warn-dangling-local.cpp b/clang/test/SemaCXX/warn-dangling-local.cpp index a946b8a2..2808a4c 100644 --- a/clang/test/SemaCXX/warn-dangling-local.cpp +++ b/clang/test/SemaCXX/warn-dangling-local.cpp @@ -26,3 +26,14 @@ void g() { const int a[] = {a[0]}; const int b[] = {a[0]}; } + +namespace std { +// std::basic_string has a hard-coded gsl::owner attr. +struct basic_string { + const char* c_str(); +}; +} // namespace std +void test(const char* a) { + // verify we're emitting the `-Wdangling-assignment` warning. + a = std::basic_string().c_str(); // expected-warning {{object backing the pointer a will be destroyed at the end of the full-expression}} +} diff --git a/clang/test/SemaTemplate/aggregate-deduction-candidate.cpp b/clang/test/SemaTemplate/aggregate-deduction-candidate.cpp index 5a50332..49afb6b 100644 --- a/clang/test/SemaTemplate/aggregate-deduction-candidate.cpp +++ b/clang/test/SemaTemplate/aggregate-deduction-candidate.cpp @@ -2,7 +2,7 @@ // RUN: %clang_cc1 -std=c++20 -verify=expected,cxx20 -ast-dump -ast-dump-decl-types -ast-dump-filter "deduction guide" %s | FileCheck %s --strict-whitespace namespace Basic { - template<class T> struct A { // cxx17-note 6 {{candidate}} + template<class T> struct A { // cxx17-note 6 {{candidate}} cxx17-note 6 {{implicit deduction guide}} T x; T y; }; @@ -36,12 +36,14 @@ namespace Basic { T y; }; - template <typename T> struct C { // cxx20-note 10 {{candidate}} cxx17-note 12 {{candidate}} + template <typename T> struct C { // cxx20-note 10 {{candidate}} cxx17-note 12 {{candidate}} \ + cxx20-note 10 {{implicit deduction guide}} cxx17-note 12 {{implicit deduction guide}} S<T> s; T t; }; - template <typename T> struct D { // cxx20-note 6 {{candidate}} cxx17-note 8 {{candidate}} + template <typename T> struct D { // cxx20-note 6 {{candidate}} cxx17-note 8 {{candidate}} \ + cxx20-note 6 {{implicit deduction guide}} cxx17-note 8 {{implicit deduction guide}} S<int> s; T t; }; @@ -99,7 +101,7 @@ namespace Basic { // CHECK: |-ClassTemplateSpecialization {{.*}} 'S' // CHECK: `-BuiltinType {{.*}} 'int' - template <typename T> struct E { // cxx17-note 4 {{candidate}} + template <typename T> struct E { // cxx17-note 4 {{candidate}} cxx17-note 4 {{implicit deduction guide}} T t; decltype(t) t2; }; @@ -133,7 +135,7 @@ namespace Basic { }; template <typename T> - struct F { // cxx17-note 2 {{candidate}} + struct F { // cxx17-note 2 {{candidate}} cxx17-note 2 {{implicit deduction guide}} typename I<T>::type i; T t; }; @@ -161,7 +163,8 @@ namespace Basic { namespace Array { typedef unsigned long size_t; - template <typename T, size_t N> struct A { // cxx20-note 2 {{candidate}} cxx17-note 14 {{candidate}} + template <typename T, size_t N> struct A { // cxx20-note 2 {{candidate}} cxx17-note 14 {{candidate}} \ + cxx20-note 2 {{implicit deduction guide}} cxx17-note 14 {{implicit deduction guide}} T array[N]; }; @@ -217,7 +220,7 @@ namespace Array { } namespace BraceElision { - template <typename T> struct A { // cxx17-note 4 {{candidate}} + template <typename T> struct A { // cxx17-note 4 {{candidate}} cxx17-note 4 {{implicit deduction guide}} T array[2]; }; @@ -247,7 +250,7 @@ namespace BraceElision { } namespace TrailingPack { - template<typename... T> struct A : T... { // cxx17-note 4 {{candidate}} + template<typename... T> struct A : T... { // cxx17-note 4 {{candidate}} cxx17-note 4 {{implicit deduction guide}} }; A a1 = { // cxx17-error {{no viable}} @@ -286,7 +289,7 @@ namespace TrailingPack { } namespace NonTrailingPack { - template<typename... T> struct A : T... { // expected-note 4 {{candidate}} + template<typename... T> struct A : T... { // expected-note 4 {{candidate}} expected-note 4 {{implicit deduction guide}} int a; }; @@ -303,7 +306,8 @@ namespace NonTrailingPack { namespace DeduceArity { template <typename... T> struct Types {}; - template <typename... T> struct F : Types<T...>, T... {}; // cxx20-note 12 {{candidate}} cxx17-note 16 {{candidate}} + template <typename... T> struct F : Types<T...>, T... {}; // cxx20-note 12 {{candidate}} cxx17-note 16 {{candidate}} \ + cxx20-note 12 {{implicit deduction guide}} cxx17-note 16 {{implicit deduction guide}} struct X {}; struct Y {}; diff --git a/clang/test/SemaTemplate/class-template-id.cpp b/clang/test/SemaTemplate/class-template-id.cpp index eade4f6..96696f6 100644 --- a/clang/test/SemaTemplate/class-template-id.cpp +++ b/clang/test/SemaTemplate/class-template-id.cpp @@ -45,6 +45,8 @@ typedef N::C<float> c2; // PR5655 template<typename T> struct Foo { }; // precxx17-note {{template is declared here}} \ cxx17-note {{candidate template ignored: couldn't infer template argument 'T'}} \ + cxx17-note {{implicit deduction guide declared as 'template <typename T> Foo() -> Foo<T>'}} \ + cxx17-note {{implicit deduction guide declared as 'template <typename T> Foo(Foo<T>) -> Foo<T>'}} \ cxx17-note {{candidate function template not viable: requires 1 argument, but 0 were provided}} void f(void) { Foo bar; } // precxx17-error {{use of class template 'Foo' requires template arguments}} \ diff --git a/clang/test/SemaTemplate/ctad.cpp b/clang/test/SemaTemplate/ctad.cpp index e981ea8..1bf605f 100644 --- a/clang/test/SemaTemplate/ctad.cpp +++ b/clang/test/SemaTemplate/ctad.cpp @@ -20,8 +20,10 @@ namespace Access { protected: struct type {}; }; - template<typename T> struct D : B { // expected-note {{not viable}} - D(T, typename T::type); // expected-note {{private member}} + template<typename T> struct D : B { // expected-note {{not viable}} \ + expected-note {{implicit deduction guide declared as 'template <typename T> D(D<T>) -> D<T>'}} + D(T, typename T::type); // expected-note {{private member}} \ + // expected-note {{implicit deduction guide declared as 'template <typename T> D(T, typename T::type) -> D<T>'}} }; D b = {B(), {}}; @@ -58,13 +60,15 @@ Y y(1); namespace NoCrashOnGettingDefaultArgLoc { template <typename> class A { - A(int = 1); // expected-note {{candidate template ignored: couldn't infer template argumen}} + A(int = 1); // expected-note {{candidate template ignored: couldn't infer template argumen}} \ + // expected-note {{implicit deduction guide declared as 'template <typename> D(int = <null expr>) -> D<type-parameter-0-0>'}} }; class C : A<int> { using A::A; }; template <typename> -class D : C { // expected-note {{candidate function template not viable: requires 1 argument}} +class D : C { // expected-note {{candidate function template not viable: requires 1 argument}} \ + expected-note {{implicit deduction guide declared as 'template <typename> D(D<type-parameter-0-0>) -> D<type-parameter-0-0>'}} using C::C; }; D abc; // expected-error {{no viable constructor or deduction guide}} diff --git a/clang/test/SemaTemplate/deduction-crash.cpp b/clang/test/SemaTemplate/deduction-crash.cpp index 86ec9f7..e3f97c5 100644 --- a/clang/test/SemaTemplate/deduction-crash.cpp +++ b/clang/test/SemaTemplate/deduction-crash.cpp @@ -166,8 +166,9 @@ namespace PR51872_part1 { template<int> class T1 { template <struct U1> T1(); }; // expected-error@-1 {{non-type template parameter has incomplete type 'struct U1'}} // expected-note@-2 {{forward declaration of 'PR51872_part1::U1'}} + // expected-note@-3 {{implicit deduction guide declared as 'template <int> T1(T1<>) -> T1<>'}} T1 t1 = 0; // expected-error@-1 {{no viable constructor or deduction guide for deduction of template arguments of 'T1'}} - // expected-note@-6 {{candidate template ignored: could not match 'T1<>' against 'int'}} + // expected-note@-7 {{candidate template ignored: could not match 'T1<>' against 'int'}} } diff --git a/clang/test/SemaTemplate/deduction-guide.cpp b/clang/test/SemaTemplate/deduction-guide.cpp index 163b365..7b9ca4d 100644 --- a/clang/test/SemaTemplate/deduction-guide.cpp +++ b/clang/test/SemaTemplate/deduction-guide.cpp @@ -123,9 +123,11 @@ using CT = C<int>; // CHECK: | `-TemplateArgument template // CHECK: `-TemplateTypeParmType {{.*}} 'type-parameter-0-2' dependent depth 0 index 2 -template<typename ...T> struct D { // expected-note {{candidate}} +template<typename ...T> struct D { // expected-note {{candidate}} \ + // expected-note {{implicit deduction guide declared as 'template <typename ...T> D(D<T...>) -> D<T...>'}} template<typename... U> using B = int(int (*...p)(T, U)); - template<typename U1, typename U2> D(B<U1, U2>*); // expected-note {{candidate}} + template<typename U1, typename U2> D(B<U1, U2>*); // expected-note {{candidate}} \ + // expected-note {{implicit deduction guide declared as 'template <typename ...T, typename U1, typename U2> D(B<type-parameter-0-1, type-parameter-0-2> *) -> D<T...>'}} }; int f(int(int, int), int(int, int)); // FIXME: We can't deduce this because we can't deduce through a @@ -166,9 +168,11 @@ using DT = D<int, int>; // CHECK-NOT: Subst // CHECK: `-TemplateTypeParmType -template<int ...N> struct E { // expected-note {{candidate}} +template<int ...N> struct E { // expected-note {{candidate}} \ + expected-note {{implicit deduction guide declared as 'template <int ...N> E(E<N...>) -> E<N...>'}} template<int ...M> using B = Z<X<N, M>...>; - template<int M1, int M2> E(B<M1, M2>); // expected-note {{candidate}} + template<int M1, int M2> E(B<M1, M2>); // expected-note {{candidate}} \ + // expected-note {{implicit deduction guide declared as 'template <int ...N, int M1, int M2> E(B<M1, M2>) -> E<N...>'}}}} }; // FIXME: We can't deduce this because we can't deduce through a // SubstNonTypeTemplateParmPackExpr. diff --git a/clang/test/SemaTemplate/nested-implicit-deduction-guides.cpp b/clang/test/SemaTemplate/nested-implicit-deduction-guides.cpp index a4ae046..af3e335 100644 --- a/clang/test/SemaTemplate/nested-implicit-deduction-guides.cpp +++ b/clang/test/SemaTemplate/nested-implicit-deduction-guides.cpp @@ -82,8 +82,11 @@ using NIL = nested_init_list<int>::B<int>; // expected-error@+1 {{no viable constructor or deduction guide for deduction of template arguments of 'nested_init_list<int>::concept_fail'}} nested_init_list<int>::concept_fail nil_invalid{1, ""}; // expected-note@#INIT_LIST_INNER_INVALID {{candidate template ignored: substitution failure [with F = const char *]: constraints not satisfied for class template 'concept_fail' [with F = const char *]}} +// expected-note@#INIT_LIST_INNER_INVALID {{implicit deduction guide declared as 'template <False F> concept_fail(int, F) -> concept_fail<F>'}} // expected-note@#INIT_LIST_INNER_INVALID {{candidate function template not viable: requires 1 argument, but 2 were provided}} +// expected-note@#INIT_LIST_INNER_INVALID {{implicit deduction guide declared as 'template <False F> concept_fail(concept_fail<F>) -> concept_fail<F>'}} // expected-note@#INIT_LIST_INNER_INVALID {{candidate function template not viable: requires 0 arguments, but 2 were provided}} +// expected-note@#INIT_LIST_INNER_INVALID {{implicit deduction guide declared as 'template <False F> concept_fail() -> concept_fail<F>'}} namespace GH88142 { diff --git a/clang/test/SemaTemplate/temp_arg.cpp b/clang/test/SemaTemplate/temp_arg.cpp index 2db4a51..538056a 100644 --- a/clang/test/SemaTemplate/temp_arg.cpp +++ b/clang/test/SemaTemplate/temp_arg.cpp @@ -6,7 +6,9 @@ template<typename T, class A; // precxx17-note 3 {{template is declared here}} \ cxx17-note 2 {{template is declared here}} \ cxx17-note {{candidate template ignored: couldn't infer template argument 'T'}} \ - cxx17-note {{candidate function template not viable: requires 1 argument, but 0 were provided}} + cxx17-note {{implicit deduction guide declared as 'template <typename T, int I, template <typename> class TT> A(A<T, I, TT>) -> A<T, I, TT>'}} \ + cxx17-note {{candidate function template not viable: requires 1 argument, but 0 were provided}} \ + cxx17-note {{implicit deduction guide declared as 'template <typename T, int I, template <typename> class TT> A() -> A<T, I, TT>'}} \ template<typename> class X; diff --git a/clang/unittests/Format/ConfigParseTest.cpp b/clang/unittests/Format/ConfigParseTest.cpp index 2466677a..cc04415 100644 --- a/clang/unittests/Format/ConfigParseTest.cpp +++ b/clang/unittests/Format/ConfigParseTest.cpp @@ -240,6 +240,7 @@ TEST(ConfigParseTest, ParsesConfigurationBools) { CHECK_PARSE_NESTED_BOOL(SpaceBeforeParensOptions, AfterOverloadedOperator); CHECK_PARSE_NESTED_BOOL(SpaceBeforeParensOptions, AfterPlacementOperator); CHECK_PARSE_NESTED_BOOL(SpaceBeforeParensOptions, BeforeNonEmptyParentheses); + CHECK_PARSE_NESTED_BOOL(SpacesInParensOptions, ExceptDoubleParentheses); CHECK_PARSE_NESTED_BOOL(SpacesInParensOptions, InCStyleCasts); CHECK_PARSE_NESTED_BOOL(SpacesInParensOptions, InConditionalStatements); CHECK_PARSE_NESTED_BOOL(SpacesInParensOptions, InEmptyParentheses); @@ -626,20 +627,24 @@ TEST(ConfigParseTest, ParsesConfiguration) { FormatStyle::SIPO_Custom); Style.SpacesInParens = FormatStyle::SIPO_Never; Style.SpacesInParensOptions = {}; - CHECK_PARSE("SpacesInParentheses: true", SpacesInParensOptions, - FormatStyle::SpacesInParensCustom(true, false, false, true)); + CHECK_PARSE( + "SpacesInParentheses: true", SpacesInParensOptions, + FormatStyle::SpacesInParensCustom(false, true, false, false, true)); Style.SpacesInParens = FormatStyle::SIPO_Never; Style.SpacesInParensOptions = {}; - CHECK_PARSE("SpacesInConditionalStatement: true", SpacesInParensOptions, - FormatStyle::SpacesInParensCustom(true, false, false, false)); + CHECK_PARSE( + "SpacesInConditionalStatement: true", SpacesInParensOptions, + FormatStyle::SpacesInParensCustom(false, true, false, false, false)); Style.SpacesInParens = FormatStyle::SIPO_Never; Style.SpacesInParensOptions = {}; - CHECK_PARSE("SpacesInCStyleCastParentheses: true", SpacesInParensOptions, - FormatStyle::SpacesInParensCustom(false, true, false, false)); + CHECK_PARSE( + "SpacesInCStyleCastParentheses: true", SpacesInParensOptions, + FormatStyle::SpacesInParensCustom(false, false, true, false, false)); Style.SpacesInParens = FormatStyle::SIPO_Never; Style.SpacesInParensOptions = {}; - CHECK_PARSE("SpaceInEmptyParentheses: true", SpacesInParensOptions, - FormatStyle::SpacesInParensCustom(false, false, true, false)); + CHECK_PARSE( + "SpaceInEmptyParentheses: true", SpacesInParensOptions, + FormatStyle::SpacesInParensCustom(false, false, false, true, false)); Style.SpacesInParens = FormatStyle::SIPO_Never; Style.SpacesInParensOptions = {}; diff --git a/clang/unittests/Format/FormatTest.cpp b/clang/unittests/Format/FormatTest.cpp index 5276e79..283843a 100644 --- a/clang/unittests/Format/FormatTest.cpp +++ b/clang/unittests/Format/FormatTest.cpp @@ -17129,6 +17129,23 @@ TEST_F(FormatTest, ConfigurableSpacesInParens) { verifyFormat("SomeType *__attribute__((attr)) *a = NULL;", Spaces); verifyFormat("void __attribute__((naked)) foo(int bar)", Spaces); verifyFormat("void f() __attribute__((asdf));", Spaces); + verifyFormat("x = (int32)y;", Spaces); + verifyFormat("y = ((int (*)(int))foo)(x);", Spaces); + verifyFormat("decltype(x) y = 42;", Spaces); + verifyFormat("decltype((x)) y = z;", Spaces); + verifyFormat("decltype((foo())) a = foo();", Spaces); + verifyFormat("decltype((bar(10))) a = bar(11);", Spaces); + verifyFormat("if ((x - y) && (a ^ b))\n" + " f();", + Spaces); + verifyFormat("for (int i = 0; i < 10; i = (i + 1))\n" + " foo(i);", + Spaces); + verifyFormat("switch (x / (y + z)) {\n" + "default:\n" + " break;\n" + "}", + Spaces); Spaces.SpacesInParens = FormatStyle::SIPO_Custom; Spaces.SpacesInParensOptions = {}; @@ -17163,6 +17180,23 @@ TEST_F(FormatTest, ConfigurableSpacesInParens) { verifyFormat("SomeType *__attribute__( ( attr ) ) *a = NULL;", Spaces); verifyFormat("void __attribute__( ( naked ) ) foo( int bar )", Spaces); verifyFormat("void f() __attribute__( ( asdf ) );", Spaces); + verifyFormat("x = (int32)y;", Spaces); + verifyFormat("y = ( (int ( * )( int ))foo )( x );", Spaces); + verifyFormat("decltype( x ) y = 42;", Spaces); + verifyFormat("decltype( ( x ) ) y = z;", Spaces); + verifyFormat("decltype( ( foo() ) ) a = foo();", Spaces); + verifyFormat("decltype( ( bar( 10 ) ) ) a = bar( 11 );", Spaces); + verifyFormat("if ( ( x - y ) && ( a ^ b ) )\n" + " f();", + Spaces); + verifyFormat("for ( int i = 0; i < 10; i = ( i + 1 ) )\n" + " foo( i );", + Spaces); + verifyFormat("switch ( x / ( y + z ) ) {\n" + "default:\n" + " break;\n" + "}", + Spaces); Spaces.SpacesInParens = FormatStyle::SIPO_Custom; Spaces.SpacesInParensOptions = {}; @@ -17175,6 +17209,7 @@ TEST_F(FormatTest, ConfigurableSpacesInParens) { verifyFormat("#define AA(X) sizeof((( X * )NULL)->a)", Spaces); verifyFormat("my_int a = ( my_int )sizeof(int);", Spaces); verifyFormat("#define x (( int )-1)", Spaces); + verifyFormat("y = (( int (*)(int) )foo)(x);", Spaces); // Run the first set of tests again with: Spaces.SpacesInParens = FormatStyle::SIPO_Custom; @@ -17207,6 +17242,23 @@ TEST_F(FormatTest, ConfigurableSpacesInParens) { verifyFormat("SomeType *__attribute__((attr)) *a = NULL;", Spaces); verifyFormat("void __attribute__((naked)) foo(int bar)", Spaces); verifyFormat("void f( ) __attribute__((asdf));", Spaces); + verifyFormat("x = ( int32 )y;", Spaces); + verifyFormat("y = (( int (*)(int) )foo)(x);", Spaces); + verifyFormat("decltype(x) y = 42;", Spaces); + verifyFormat("decltype((x)) y = z;", Spaces); + verifyFormat("decltype((foo( ))) a = foo( );", Spaces); + verifyFormat("decltype((bar(10))) a = bar(11);", Spaces); + verifyFormat("if ((x - y) && (a ^ b))\n" + " f( );", + Spaces); + verifyFormat("for (int i = 0; i < 10; i = (i + 1))\n" + " foo(i);", + Spaces); + verifyFormat("switch (x / (y + z)) {\n" + "default:\n" + " break;\n" + "}", + Spaces); // Run the first set of tests again with: Spaces.SpaceAfterCStyleCast = true; @@ -17314,6 +17366,63 @@ TEST_F(FormatTest, ConfigurableSpacesInParens) { verifyFormat("size_t idx = (a->foo)(a - 1);", Spaces); verifyFormat("size_t idx = (*foo)(a - 1);", Spaces); verifyFormat("size_t idx = (*(foo))(a - 1);", Spaces); + + // Check ExceptDoubleParentheses spaces + Spaces.IndentWidth = 2; + Spaces.SpacesInParens = FormatStyle::SIPO_Custom; + Spaces.SpacesInParensOptions = {}; + Spaces.SpacesInParensOptions.Other = true; + Spaces.SpacesInParensOptions.ExceptDoubleParentheses = true; + verifyFormat("SomeType *__attribute__(( attr )) *a = NULL;", Spaces); + verifyFormat("void __attribute__(( naked )) foo( int bar )", Spaces); + verifyFormat("void f() __attribute__(( asdf ));", Spaces); + verifyFormat("__attribute__(( __aligned__( x ) )) z;", Spaces); + verifyFormat("int x __attribute__(( aligned( 16 ) )) = 0;", Spaces); + verifyFormat("class __declspec( dllimport ) X {};", Spaces); + verifyFormat("class __declspec(( dllimport )) X {};", Spaces); + verifyFormat("int x = ( ( a - 1 ) * 3 );", Spaces); + verifyFormat("int x = ( 3 * ( a - 1 ) );", Spaces); + verifyFormat("decltype( x ) y = 42;", Spaces); + verifyFormat("decltype(( bar( 10 ) )) a = bar( 11 );", Spaces); + verifyFormat("if (( i = j ))\n" + " do_something( i );", + Spaces); + + Spaces.SpacesInParens = FormatStyle::SIPO_Custom; + Spaces.SpacesInParensOptions = {}; + Spaces.SpacesInParensOptions.InConditionalStatements = true; + Spaces.SpacesInParensOptions.ExceptDoubleParentheses = true; + verifyFormat("while ( (bool)1 )\n" + " continue;", + Spaces); + verifyFormat("while ((i = j))\n" + " continue;", + Spaces); + verifyFormat("do {\n" + " do_something((int)i);\n" + "} while ( something() );", + Spaces); + verifyFormat("do {\n" + " do_something((int)i);\n" + "} while ((i = i + 1));", + Spaces); + verifyFormat("if ( (x - y) && (a ^ b) )\n" + " f();", + Spaces); + verifyFormat("if ((i = j))\n" + " do_something(i);", + Spaces); + verifyFormat("for ( int i = 0; i < 10; i = (i + 1) )\n" + " foo(i);", + Spaces); + verifyFormat("switch ( x / (y + z) ) {\n" + "default:\n" + " break;\n" + "}", + Spaces); + verifyFormat("if constexpr ((a = b))\n" + " c;", + Spaces); } TEST_F(FormatTest, ConfigurableSpacesInSquareBrackets) { diff --git a/clang/www/c_status.html b/clang/www/c_status.html index 8e42b81..ccb39a1 100644 --- a/clang/www/c_status.html +++ b/clang/www/c_status.html @@ -71,7 +71,7 @@ </tr> <tr> <td><a href="#c2y">C2y</a></td> - <td><tt>(Flag currently unavailable)</tt></td> + <td><tt>-std=c2y</tt></td> <td class="partial" align="center">Partial</td> </tr> </table> @@ -1229,8 +1229,7 @@ conforms by not defining the <code>__STDC_IEC_559_COMPLEX__</code> macro. <p>Clang has support for some of the features of the C standard following C23, informally referred to as C2y.</p> -<p>Clang currently does not expose a language standard mode flag for C2y. -<!--You can use Clang in C2y mode with the <code>-std=c2y</code> option (available in Clang 19 and later).--></p> +<p>You can use Clang in C2y mode with the <code>-std=c2y</code> option (available in Clang 19 and later).</p> <details open> <summary>List of features and minimum Clang version with support</summary> diff --git a/compiler-rt/lib/asan/asan_suppressions.cpp b/compiler-rt/lib/asan/asan_suppressions.cpp index e71d231..6cee674 100644 --- a/compiler-rt/lib/asan/asan_suppressions.cpp +++ b/compiler-rt/lib/asan/asan_suppressions.cpp @@ -39,8 +39,7 @@ void InitializeSuppressions() { suppression_ctx = new (suppression_placeholder) SuppressionContext(kSuppressionTypes, ARRAY_SIZE(kSuppressionTypes)); suppression_ctx->ParseFromFile(flags()->suppressions); - if (&__asan_default_suppressions) - suppression_ctx->Parse(__asan_default_suppressions()); + suppression_ctx->Parse(__asan_default_suppressions()); } bool IsInterceptorSuppressed(const char *interceptor_name) { diff --git a/compiler-rt/lib/sanitizer_common/sancov_flags.cpp b/compiler-rt/lib/sanitizer_common/sancov_flags.cpp index ed46e88..43b4bcc 100644 --- a/compiler-rt/lib/sanitizer_common/sancov_flags.cpp +++ b/compiler-rt/lib/sanitizer_common/sancov_flags.cpp @@ -37,10 +37,6 @@ static void RegisterSancovFlags(FlagParser *parser, SancovFlags *f) { #undef SANCOV_FLAG } -static const char *MaybeCallSancovDefaultOptions() { - return (&__sancov_default_options) ? __sancov_default_options() : ""; -} - void InitializeSancovFlags() { SancovFlags *f = sancov_flags(); f->SetDefaults(); @@ -48,7 +44,7 @@ void InitializeSancovFlags() { FlagParser parser; RegisterSancovFlags(&parser, f); - parser.ParseString(MaybeCallSancovDefaultOptions()); + parser.ParseString(__sancov_default_options()); parser.ParseStringFromEnv("SANCOV_OPTIONS"); ReportUnrecognizedFlags(); diff --git a/flang/include/flang/Optimizer/Builder/FIRBuilder.h b/flang/include/flang/Optimizer/Builder/FIRBuilder.h index f9ef8b7..ea35b29 100644 --- a/flang/include/flang/Optimizer/Builder/FIRBuilder.h +++ b/flang/include/flang/Optimizer/Builder/FIRBuilder.h @@ -286,6 +286,10 @@ public: fir::StringLitOp createStringLitOp(mlir::Location loc, llvm::StringRef string); + std::pair<fir::TypeInfoOp, mlir::OpBuilder::InsertPoint> + createTypeInfoOp(mlir::Location loc, fir::RecordType recordType, + fir::RecordType parentType); + //===--------------------------------------------------------------------===// // Linkage helpers (inline). The default linkage is external. //===--------------------------------------------------------------------===// diff --git a/flang/include/flang/Optimizer/CodeGen/FIROpPatterns.h b/flang/include/flang/Optimizer/CodeGen/FIROpPatterns.h index ac09566..91b20251 100644 --- a/flang/include/flang/Optimizer/CodeGen/FIROpPatterns.h +++ b/flang/include/flang/Optimizer/CodeGen/FIROpPatterns.h @@ -205,10 +205,10 @@ public: rewrite(mlir::cast<SourceOp>(op), OpAdaptor(operands, mlir::cast<SourceOp>(op)), rewriter); } - mlir::LogicalResult match(mlir::Operation *op) const final { + llvm::LogicalResult match(mlir::Operation *op) const final { return match(mlir::cast<SourceOp>(op)); } - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::Operation *op, mlir::ArrayRef<mlir::Value> operands, mlir::ConversionPatternRewriter &rewriter) const final { return matchAndRewrite(mlir::cast<SourceOp>(op), @@ -218,14 +218,14 @@ public: /// Rewrite and Match methods that operate on the SourceOp type. These must be /// overridden by the derived pattern class. - virtual mlir::LogicalResult match(SourceOp op) const { + virtual llvm::LogicalResult match(SourceOp op) const { llvm_unreachable("must override match or matchAndRewrite"); } virtual void rewrite(SourceOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const { llvm_unreachable("must override rewrite or matchAndRewrite"); } - virtual mlir::LogicalResult + virtual llvm::LogicalResult matchAndRewrite(SourceOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const { if (mlir::failed(match(op))) @@ -246,14 +246,14 @@ public: using FIROpConversion<FromOp>::FIROpConversion; using OpAdaptor = typename FromOp::Adaptor; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(FromOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const final { mlir::Type ty = this->convertType(op.getType()); return doRewrite(op, ty, adaptor, rewriter); } - virtual mlir::LogicalResult + virtual llvm::LogicalResult doRewrite(FromOp addr, mlir::Type ty, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const = 0; }; diff --git a/flang/include/flang/Optimizer/CodeGen/TypeConverter.h b/flang/include/flang/Optimizer/CodeGen/TypeConverter.h index 58803a5..ece3e4e 100644 --- a/flang/include/flang/Optimizer/CodeGen/TypeConverter.h +++ b/flang/include/flang/Optimizer/CodeGen/TypeConverter.h @@ -60,7 +60,7 @@ public: mlir::Type indexType() const; // fir.type<name(p : TY'...){f : TY...}> --> llvm<"%name = { ty... }"> - std::optional<mlir::LogicalResult> + std::optional<llvm::LogicalResult> convertRecordType(fir::RecordType derived, llvm::SmallVectorImpl<mlir::Type> &results); diff --git a/flang/include/flang/Optimizer/Dialect/FIROps.td b/flang/include/flang/Optimizer/Dialect/FIROps.td index baf0952..5b03806 100644 --- a/flang/include/flang/Optimizer/Dialect/FIROps.td +++ b/flang/include/flang/Optimizer/Dialect/FIROps.td @@ -2956,7 +2956,10 @@ def fir_TypeInfoOp : fir_Op<"type_info", let hasVerifier = 1; - let regions = (region MaxSizedRegion<1>:$dispatch_table); + let regions = (region + MaxSizedRegion<1>:$dispatch_table, + MaxSizedRegion<1>:$component_info + ); let builders = [ OpBuilder<(ins "fir::RecordType":$type, "fir::RecordType":$parent_type, @@ -2967,6 +2970,7 @@ def fir_TypeInfoOp : fir_Op<"type_info", $sym_name (`noinit` $no_init^)? (`nodestroy` $no_destroy^)? (`nofinal` $no_final^)? (`extends` $parent_type^)? attr-dict `:` $type (`dispatch_table` $dispatch_table^)? + (`component_info` $component_info^)? }]; let extraClassDeclaration = [{ @@ -3010,6 +3014,24 @@ def fir_DTEntryOp : fir_Op<"dt_entry", [HasParent<"TypeInfoOp">]> { }]; } +def fir_DTComponentOp : fir_Op<"dt_component", [HasParent<"TypeInfoOp">]> { + let summary = "define extra information about a component inside fir.type_info"; + + let description = [{ + ``` + fir.dt_component i lbs [-1,2] init @init_val + ``` + }]; + + let arguments = (ins + StrAttr:$name, + OptionalAttr<DenseI64ArrayAttr>:$lower_bounds, + OptionalAttr<FlatSymbolRefAttr>:$init_val + ); + + let assemblyFormat = "$name (`lbs` $lower_bounds^)? (`init` $init_val^)? attr-dict"; +} + def fir_AbsentOp : fir_OneResultOp<"absent", [NoMemoryEffect]> { let summary = "create value to be passed for absent optional function argument"; let description = [{ diff --git a/flang/include/flang/Optimizer/Dialect/FirAliasTagOpInterface.h b/flang/include/flang/Optimizer/Dialect/FirAliasTagOpInterface.h index f2d5b39..1e3728e 100644 --- a/flang/include/flang/Optimizer/Dialect/FirAliasTagOpInterface.h +++ b/flang/include/flang/Optimizer/Dialect/FirAliasTagOpInterface.h @@ -16,10 +16,10 @@ #include "mlir/IR/OpDefinition.h" #include "mlir/IR/Operation.h" -#include "mlir/Support/LogicalResult.h" +#include "llvm/Support/LogicalResult.h" namespace fir::detail { -mlir::LogicalResult verifyFirAliasTagOpInterface(mlir::Operation *op); +llvm::LogicalResult verifyFirAliasTagOpInterface(mlir::Operation *op); } // namespace fir::detail #include "flang/Optimizer/Dialect/FirAliasTagOpInterface.h.inc" diff --git a/flang/include/flang/Optimizer/Dialect/FortranVariableInterface.td b/flang/include/flang/Optimizer/Dialect/FortranVariableInterface.td index 3f78a93..926e00c 100644 --- a/flang/include/flang/Optimizer/Dialect/FortranVariableInterface.td +++ b/flang/include/flang/Optimizer/Dialect/FortranVariableInterface.td @@ -192,7 +192,7 @@ def fir_FortranVariableOpInterface : OpInterface<"FortranVariableOpInterface"> { } /// Interface verifier imlementation for declare operations. - mlir::LogicalResult verifyDeclareLikeOpImpl(mlir::Value memRef); + llvm::LogicalResult verifyDeclareLikeOpImpl(mlir::Value memRef); }]; diff --git a/flang/include/flang/Optimizer/Dialect/Support/KindMapping.h b/flang/include/flang/Optimizer/Dialect/Support/KindMapping.h index 138123b..083526f 100644 --- a/flang/include/flang/Optimizer/Dialect/Support/KindMapping.h +++ b/flang/include/flang/Optimizer/Dialect/Support/KindMapping.h @@ -124,7 +124,7 @@ public: private: MatchResult badMapString(const llvm::Twine &ptr); MatchResult parse(llvm::StringRef kindMap); - mlir::LogicalResult setDefaultKinds(llvm::ArrayRef<KindTy> defs); + llvm::LogicalResult setDefaultKinds(llvm::ArrayRef<KindTy> defs); mlir::MLIRContext *context; llvm::DenseMap<std::pair<char, KindTy>, Bitsize> intMap; diff --git a/flang/include/flang/Optimizer/HLFIR/HLFIROps.td b/flang/include/flang/Optimizer/HLFIR/HLFIROps.td index e9915e89..fdf0db9 100644 --- a/flang/include/flang/Optimizer/HLFIR/HLFIROps.td +++ b/flang/include/flang/Optimizer/HLFIR/HLFIROps.td @@ -733,7 +733,7 @@ def hlfir_AssociateOp : hlfir_Op<"associate", [AttrSizedOperandSegments, CArg<"mlir::Value", "{}">:$shape, CArg<"mlir::ValueRange", "{}">:$typeparams, CArg<"fir::FortranVariableFlagsAttr", "{}">:$fortran_attrs)>, OpBuilder<(ins "mlir::Value":$memref, CArg<"mlir::Value", "{}">:$shape, - CArg<"mlir::ValueRange", "{}">:$typeparams, + CArg<"mlir::ValueRange", "{}">:$typeparams, CArg<"fir::FortranVariableFlagsAttr", "{}">:$fortran_attrs, CArg<"llvm::ArrayRef<mlir::NamedAttribute>", "{}">:$attributes)>]; @@ -1258,7 +1258,7 @@ def hlfir_OrderedAssignmentTreeOpInterface : OpInterface<"OrderedAssignmentTreeO let extraClassDeclaration = [{ /// Interface verifier imlementation. - mlir::LogicalResult verifyImpl(); + llvm::LogicalResult verifyImpl(); mlir::Block* getSubTreeBlock() { mlir::Region* region = getSubTreeRegion(); @@ -1733,8 +1733,8 @@ def hlfir_CharExtremumOp : hlfir_Op<"char_extremum", [DeclareOpInterfaceMethods<MemoryEffectsOpInterface>]> { let summary = "Find max/min from given character strings"; let description = [{ - Find the lexicographical minimum or maximum of two or more character - strings of the same character kind and return the string with the lexicographical + Find the lexicographical minimum or maximum of two or more character + strings of the same character kind and return the string with the lexicographical minimum or maximum number of characters. Example: %0 = hlfir.char_extremum min, %arg0, %arg1 : (!fir.ref<!fir.char<1,10>>, !fir.ref<!fir.char<1,20>>) -> !hlfir.expr<!fir.char<1,10>> diff --git a/flang/include/flang/Optimizer/Support/InternalNames.h b/flang/include/flang/Optimizer/Support/InternalNames.h index 23a0385..ff23510 100644 --- a/flang/include/flang/Optimizer/Support/InternalNames.h +++ b/flang/include/flang/Optimizer/Support/InternalNames.h @@ -15,6 +15,7 @@ #include <optional> static constexpr llvm::StringRef typeDescriptorSeparator = ".dt."; +static constexpr llvm::StringRef componentInitSeparator = ".di."; static constexpr llvm::StringRef bindingTableSeparator = ".v."; static constexpr llvm::StringRef boxprocSuffix = "UnboxProc"; @@ -156,6 +157,11 @@ struct NameUniquer { static std::string getTypeDescriptorBindingTableName(llvm::StringRef mangledTypeName); + /// Given a mangled derived type name and a component name, get the name of + /// the global object containing the component default initialization. + static std::string getComponentInitName(llvm::StringRef mangledTypeName, + llvm::StringRef componentName); + /// Remove markers that have been added when doing partial type /// conversions. mlir::Type cannot be mutated in a pass, so new /// fir::RecordType must be created when lowering member types. diff --git a/flang/include/flang/Optimizer/Support/Utils.h b/flang/include/flang/Optimizer/Support/Utils.h index d8bcb5f..ae95a26 100644 --- a/flang/include/flang/Optimizer/Support/Utils.h +++ b/flang/include/flang/Optimizer/Support/Utils.h @@ -137,6 +137,29 @@ inline void intrinsicTypeTODO(fir::FirOpBuilder &builder, mlir::Type type, " in " + intrinsicName); } +/// Find the fir.type_info that was created for this \p recordType in \p module, +/// if any. \p symbolTable can be provided to speed-up the lookup. This tool +/// will match record type even if they have been "altered" in type conversion +/// passes. +fir::TypeInfoOp +lookupTypeInfoOp(fir::RecordType recordType, mlir::ModuleOp module, + const mlir::SymbolTable *symbolTable = nullptr); + +/// Find the fir.type_info named \p name in \p module, if any. \p symbolTable +/// can be provided to speed-up the lookup. Prefer using the equivalent with a +/// RecordType argument unless it is certain \p name has not been altered by a +/// pass rewriting fir.type (see NameUniquer::dropTypeConversionMarkers). +fir::TypeInfoOp +lookupTypeInfoOp(llvm::StringRef name, mlir::ModuleOp module, + const mlir::SymbolTable *symbolTable = nullptr); + +/// Returns all lower bounds of \p component if it is an array component of \p +/// recordType with non default lower bounds. Returns nullopt if this is not an +/// array componnet of \p recordType or if its lower bounds are all ones. +std::optional<llvm::ArrayRef<int64_t>> getComponentLowerBoundsIfNonDefault( + fir::RecordType recordType, llvm::StringRef component, + mlir::ModuleOp module, const mlir::SymbolTable *symbolTable = nullptr); + } // namespace fir #endif // FORTRAN_OPTIMIZER_SUPPORT_UTILS_H diff --git a/flang/lib/Lower/Bridge.cpp b/flang/lib/Lower/Bridge.cpp index d4c5053..60422dd 100644 --- a/flang/lib/Lower/Bridge.cpp +++ b/flang/lib/Lower/Bridge.cpp @@ -148,6 +148,71 @@ struct ConstructContext { bool pushedScope = false; // was a scoped pushed for this construct? }; +/// Helper to gather the lower bounds of array components with non deferred +/// shape when they are not all ones. Return an empty array attribute otherwise. +static mlir::DenseI64ArrayAttr +gatherComponentNonDefaultLowerBounds(mlir::Location loc, + mlir::MLIRContext *mlirContext, + const Fortran::semantics::Symbol &sym) { + if (Fortran::semantics::IsAllocatableOrObjectPointer(&sym)) + return {}; + mlir::DenseI64ArrayAttr lbs_attr; + if (const auto *objDetails = + sym.detailsIf<Fortran::semantics::ObjectEntityDetails>()) { + llvm::SmallVector<std::int64_t> lbs; + bool hasNonDefaultLbs = false; + for (const Fortran::semantics::ShapeSpec &bounds : objDetails->shape()) + if (auto lb = bounds.lbound().GetExplicit()) { + if (auto constant = Fortran::evaluate::ToInt64(*lb)) { + hasNonDefaultLbs |= (*constant != 1); + lbs.push_back(*constant); + } else { + TODO(loc, "generate fir.dt_component for length parametrized derived " + "types"); + } + } + if (hasNonDefaultLbs) { + assert(static_cast<int>(lbs.size()) == sym.Rank() && + "expected component bounds to be constant or deferred"); + lbs_attr = mlir::DenseI64ArrayAttr::get(mlirContext, lbs); + } + } + return lbs_attr; +} + +// Helper class to generate name of fir.global containing component explicit +// default value for objects, and initial procedure target for procedure pointer +// components. +static mlir::FlatSymbolRefAttr gatherComponentInit( + mlir::Location loc, Fortran::lower::AbstractConverter &converter, + const Fortran::semantics::Symbol &sym, fir::RecordType derivedType) { + mlir::MLIRContext *mlirContext = &converter.getMLIRContext(); + // Return procedure target mangled name for procedure pointer components. + if (const auto *procPtr = + sym.detailsIf<Fortran::semantics::ProcEntityDetails>()) { + if (std::optional<const Fortran::semantics::Symbol *> maybeInitSym = + procPtr->init()) { + // So far, do not make distinction between p => NULL() and p without init, + // f18 always initialize pointers to NULL anyway. + if (!*maybeInitSym) + return {}; + return mlir::FlatSymbolRefAttr::get(mlirContext, + converter.mangleName(**maybeInitSym)); + } + } + + const auto *objDetails = + sym.detailsIf<Fortran::semantics::ObjectEntityDetails>(); + if (!objDetails || !objDetails->init().has_value()) + return {}; + // Object component initial value. Semantic package component object default + // value into compiler generated symbols that are lowered as read-only + // fir.global. Get the name of this global. + std::string name = fir::NameUniquer::getComponentInitName( + derivedType.getName(), toStringRef(sym.name())); + return mlir::FlatSymbolRefAttr::get(mlirContext, name); +} + /// Helper class to generate the runtime type info global data and the /// fir.type_info operations that contain the dipatch tables (if any). /// The type info global data is required to describe the derived type to the @@ -213,15 +278,14 @@ private: parentType = mlir::cast<fir::RecordType>(converter.genType(*parent)); fir::FirOpBuilder &builder = converter.getFirOpBuilder(); - mlir::ModuleOp module = builder.getModule(); - fir::TypeInfoOp dt = - module.lookupSymbol<fir::TypeInfoOp>(info.type.getName()); - if (dt) - return; // Already created. - auto insertPt = builder.saveInsertionPoint(); - builder.setInsertionPoint(module.getBody(), module.getBody()->end()); - dt = builder.create<fir::TypeInfoOp>(info.loc, info.type, parentType); - + fir::TypeInfoOp dt; + mlir::OpBuilder::InsertPoint insertPointIfCreated; + std::tie(dt, insertPointIfCreated) = + builder.createTypeInfoOp(info.loc, info.type, parentType); + if (!insertPointIfCreated.isSet()) + return; // fir.type_info was already built in a previous call. + + // Set init, destroy, and nofinal attributes. if (!info.typeSpec.HasDefaultInitialization(/*ignoreAllocatable=*/false, /*ignorePointer=*/false)) dt->setAttr(dt.getNoInitAttrName(), builder.getUnitAttr()); @@ -230,13 +294,12 @@ private: if (!Fortran::semantics::MayRequireFinalization(info.typeSpec)) dt->setAttr(dt.getNoFinalAttrName(), builder.getUnitAttr()); - const Fortran::semantics::Scope *scope = info.typeSpec.scope(); - if (!scope) - scope = info.typeSpec.typeSymbol().scope(); - assert(scope && "failed to find type scope"); + const Fortran::semantics::Scope &derivedScope = + DEREF(info.typeSpec.GetScope()); + // Fill binding table region if the derived type has bindings. Fortran::semantics::SymbolVector bindings = - Fortran::semantics::CollectBindings(*scope); + Fortran::semantics::CollectBindings(derivedScope); if (!bindings.empty()) { builder.createBlock(&dt.getDispatchTable()); for (const Fortran::semantics::SymbolRef &binding : bindings) { @@ -252,7 +315,33 @@ private: } builder.create<fir::FirEndOp>(info.loc); } - builder.restoreInsertionPoint(insertPt); + // Gather info about components that is not reflected in fir.type and may be + // needed later: component initial values and array component non default + // lower bounds. + mlir::Block *componentInfo = nullptr; + for (const auto &componentName : + info.typeSpec.typeSymbol() + .get<Fortran::semantics::DerivedTypeDetails>() + .componentNames()) { + auto scopeIter = derivedScope.find(componentName); + assert(scopeIter != derivedScope.cend() && + "failed to find derived type component symbol"); + const Fortran::semantics::Symbol &component = scopeIter->second.get(); + mlir::FlatSymbolRefAttr init_val = + gatherComponentInit(info.loc, converter, component, info.type); + mlir::DenseI64ArrayAttr lbs = gatherComponentNonDefaultLowerBounds( + info.loc, builder.getContext(), component); + if (init_val || lbs) { + if (!componentInfo) + componentInfo = builder.createBlock(&dt.getComponentInfo()); + auto compName = mlir::StringAttr::get(builder.getContext(), + toStringRef(component.name())); + builder.create<fir::DTComponentOp>(info.loc, compName, lbs, init_val); + } + } + if (componentInfo) + builder.create<fir::FirEndOp>(info.loc); + builder.restoreInsertionPoint(insertPointIfCreated); } /// Store the front-end data that will be required to generate the type info diff --git a/flang/lib/Lower/ConvertProcedureDesignator.cpp b/flang/lib/Lower/ConvertProcedureDesignator.cpp index aa0d7ce..b528544 100644 --- a/flang/lib/Lower/ConvertProcedureDesignator.cpp +++ b/flang/lib/Lower/ConvertProcedureDesignator.cpp @@ -89,9 +89,15 @@ fir::ExtendedValue Fortran::lower::convertProcedureDesignator( fir::factory::genMaxWithZero(builder, loc, rawLen); } } + // The caller of the function pointer will have to allocate + // the function result with the character length specified + // by the boxed value. If the result length cannot be + // computed statically, set it to zero (we used to use -1, + // but this could cause assertions in LLVM after inlining + // exposed alloca of size -1). if (!funcPtrResultLength) funcPtrResultLength = builder.createIntegerConstant( - loc, builder.getCharacterLengthType(), -1); + loc, builder.getCharacterLengthType(), 0); return fir::CharBoxValue{funcPtr, funcPtrResultLength}; } return funcPtr; diff --git a/flang/lib/Optimizer/Builder/FIRBuilder.cpp b/flang/lib/Optimizer/Builder/FIRBuilder.cpp index 4a1772a..2ea302d 100644 --- a/flang/lib/Optimizer/Builder/FIRBuilder.cpp +++ b/flang/lib/Optimizer/Builder/FIRBuilder.cpp @@ -19,6 +19,7 @@ #include "flang/Optimizer/Dialect/FIRType.h" #include "flang/Optimizer/Support/FatalError.h" #include "flang/Optimizer/Support/InternalNames.h" +#include "flang/Optimizer/Support/Utils.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/OpenACC/OpenACC.h" #include "mlir/Dialect/OpenMP/OpenMPDialect.h" @@ -364,6 +365,22 @@ fir::GlobalOp fir::FirOpBuilder::createGlobal( return glob; } +std::pair<fir::TypeInfoOp, mlir::OpBuilder::InsertPoint> +fir::FirOpBuilder::createTypeInfoOp(mlir::Location loc, + fir::RecordType recordType, + fir::RecordType parentType) { + mlir::ModuleOp module = getModule(); + if (fir::TypeInfoOp typeInfo = + fir::lookupTypeInfoOp(recordType.getName(), module, symbolTable)) + return {typeInfo, InsertPoint{}}; + InsertPoint insertPoint = saveInsertionPoint(); + setInsertionPoint(module.getBody(), module.getBody()->end()); + auto typeInfo = create<fir::TypeInfoOp>(loc, recordType, parentType); + if (symbolTable) + symbolTable->insert(typeInfo); + return {typeInfo, insertPoint}; +} + mlir::Value fir::FirOpBuilder::convertWithSemantics( mlir::Location loc, mlir::Type toTy, mlir::Value val, bool allowCharacterConversion, bool allowRebox) { diff --git a/flang/lib/Optimizer/CodeGen/CodeGen.cpp b/flang/lib/Optimizer/CodeGen/CodeGen.cpp index 5f35825..7483acf 100644 --- a/flang/lib/Optimizer/CodeGen/CodeGen.cpp +++ b/flang/lib/Optimizer/CodeGen/CodeGen.cpp @@ -113,7 +113,7 @@ namespace { struct AddrOfOpConversion : public fir::FIROpConversion<fir::AddrOfOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AddrOfOp addr, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { auto ty = convertType(addr.getType()); @@ -174,7 +174,7 @@ namespace { struct DeclareOpConversion : public fir::FIROpConversion<fir::cg::XDeclareOp> { public: using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::cg::XDeclareOp declareOp, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { auto memRef = adaptor.getOperands()[0]; @@ -197,7 +197,7 @@ namespace { struct AllocaOpConversion : public fir::FIROpConversion<fir::AllocaOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AllocaOp alloc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -294,7 +294,7 @@ namespace { struct BoxAddrOpConversion : public fir::FIROpConversion<fir::BoxAddrOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxAddrOp boxaddr, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value a = adaptor.getOperands()[0]; @@ -316,7 +316,7 @@ struct BoxAddrOpConversion : public fir::FIROpConversion<fir::BoxAddrOp> { struct BoxCharLenOpConversion : public fir::FIROpConversion<fir::BoxCharLenOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxCharLenOp boxCharLen, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value boxChar = adaptor.getOperands()[0]; @@ -339,7 +339,7 @@ struct BoxCharLenOpConversion : public fir::FIROpConversion<fir::BoxCharLenOp> { struct BoxDimsOpConversion : public fir::FIROpConversion<fir::BoxDimsOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxDimsOp boxdims, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { llvm::SmallVector<mlir::Type, 3> resultTypes = { @@ -361,7 +361,7 @@ struct BoxDimsOpConversion : public fir::FIROpConversion<fir::BoxDimsOp> { struct BoxEleSizeOpConversion : public fir::FIROpConversion<fir::BoxEleSizeOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxEleSizeOp boxelesz, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value box = adaptor.getOperands()[0]; @@ -379,7 +379,7 @@ struct BoxEleSizeOpConversion : public fir::FIROpConversion<fir::BoxEleSizeOp> { struct BoxIsAllocOpConversion : public fir::FIROpConversion<fir::BoxIsAllocOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxIsAllocOp boxisalloc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value box = adaptor.getOperands()[0]; @@ -397,7 +397,7 @@ struct BoxIsAllocOpConversion : public fir::FIROpConversion<fir::BoxIsAllocOp> { struct BoxIsArrayOpConversion : public fir::FIROpConversion<fir::BoxIsArrayOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxIsArrayOp boxisarray, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value a = adaptor.getOperands()[0]; @@ -416,7 +416,7 @@ struct BoxIsArrayOpConversion : public fir::FIROpConversion<fir::BoxIsArrayOp> { struct BoxIsPtrOpConversion : public fir::FIROpConversion<fir::BoxIsPtrOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxIsPtrOp boxisptr, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value box = adaptor.getOperands()[0]; @@ -434,7 +434,7 @@ struct BoxIsPtrOpConversion : public fir::FIROpConversion<fir::BoxIsPtrOp> { struct BoxRankOpConversion : public fir::FIROpConversion<fir::BoxRankOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxRankOp boxrank, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value a = adaptor.getOperands()[0]; @@ -456,7 +456,7 @@ struct BoxProcHostOpConversion : public fir::FIROpConversion<fir::BoxProcHostOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxProcHostOp boxprochost, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { TODO(boxprochost.getLoc(), "fir.boxproc_host codegen"); @@ -470,7 +470,7 @@ struct BoxTypeDescOpConversion : public fir::FIROpConversion<fir::BoxTypeDescOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxTypeDescOp boxtypedesc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value box = adaptor.getOperands()[0]; @@ -488,7 +488,7 @@ struct BoxTypeCodeOpConversion : public fir::FIROpConversion<fir::BoxTypeCodeOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxTypeCodeOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Value box = adaptor.getOperands()[0]; @@ -506,7 +506,7 @@ struct BoxTypeCodeOpConversion struct StringLitOpConversion : public fir::FIROpConversion<fir::StringLitOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::StringLitOp constop, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { auto ty = convertType(constop.getType()); @@ -547,7 +547,7 @@ struct StringLitOpConversion : public fir::FIROpConversion<fir::StringLitOp> { struct CallOpConversion : public fir::FIROpConversion<fir::CallOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::CallOp call, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { llvm::SmallVector<mlir::Type> resultTys; @@ -578,7 +578,7 @@ namespace { struct CmpcOpConversion : public fir::FIROpConversion<fir::CmpcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::CmpcOp cmp, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -616,7 +616,7 @@ struct CmpcOpConversion : public fir::FIROpConversion<fir::CmpcOp> { struct ConstcOpConversion : public fir::FIROpConversion<fir::ConstcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ConstcOp conc, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = conc.getLoc(); @@ -647,7 +647,7 @@ struct ConvertOpConversion : public fir::FIROpConversion<fir::ConvertOp> { return mlir::isa<mlir::FloatType>(ty); } - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ConvertOp convert, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { auto fromFirTy = convert.getValue().getType(); @@ -814,7 +814,7 @@ struct ConvertOpConversion : public fir::FIROpConversion<fir::ConvertOp> { struct TypeInfoOpConversion : public fir::FIROpConversion<fir::TypeInfoOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::TypeInfoOp op, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { rewriter.eraseOp(op); @@ -827,7 +827,7 @@ struct TypeInfoOpConversion : public fir::FIROpConversion<fir::TypeInfoOp> { struct DTEntryOpConversion : public fir::FIROpConversion<fir::DTEntryOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::DTEntryOp op, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { rewriter.eraseOp(op); @@ -839,7 +839,7 @@ struct DTEntryOpConversion : public fir::FIROpConversion<fir::DTEntryOp> { struct GlobalLenOpConversion : public fir::FIROpConversion<fir::GlobalLenOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::GlobalLenOp globalLen, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { TODO(globalLen.getLoc(), "fir.global_len codegen"); @@ -853,7 +853,7 @@ struct LenParamIndexOpConversion using FIROpConversion::FIROpConversion; // FIXME: this should be specialized by the runtime target - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::LenParamIndexOp lenp, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { TODO(lenp.getLoc(), "fir.len_param_index codegen"); @@ -867,7 +867,7 @@ struct LenParamIndexOpConversion struct EmboxCharOpConversion : public fir::FIROpConversion<fir::EmboxCharOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::EmboxCharOp emboxChar, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -960,7 +960,7 @@ namespace { struct AllocMemOpConversion : public fir::FIROpConversion<fir::AllocMemOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AllocMemOp heap, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Type heapTy = heap.getType(); @@ -1030,7 +1030,7 @@ namespace { struct FreeMemOpConversion : public fir::FIROpConversion<fir::FreeMemOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::FreeMemOp freemem, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = freemem.getLoc(); @@ -1481,7 +1481,7 @@ computeTripletExtent(mlir::ConversionPatternRewriter &rewriter, struct EmboxOpConversion : public EmboxCommonConversion<fir::EmboxOp> { using EmboxCommonConversion::EmboxCommonConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::EmboxOp embox, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -1513,7 +1513,7 @@ struct EmboxOpConversion : public EmboxCommonConversion<fir::EmboxOp> { struct XEmboxOpConversion : public EmboxCommonConversion<fir::cg::XEmboxOp> { using EmboxCommonConversion::EmboxCommonConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::cg::XEmboxOp xbox, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -1711,7 +1711,7 @@ struct XEmboxOpConversion : public EmboxCommonConversion<fir::cg::XEmboxOp> { struct XReboxOpConversion : public EmboxCommonConversion<fir::cg::XReboxOp> { using EmboxCommonConversion::EmboxCommonConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::cg::XReboxOp rebox, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = rebox.getLoc(); @@ -1794,7 +1794,7 @@ struct XReboxOpConversion : public EmboxCommonConversion<fir::cg::XReboxOp> { private: /// Write resulting shape and base address in descriptor, and replace rebox /// op. - mlir::LogicalResult + llvm::LogicalResult finalizeRebox(fir::cg::XReboxOp rebox, mlir::Type destBoxTy, mlir::Value dest, mlir::Value base, mlir::ValueRange lbounds, mlir::ValueRange extents, mlir::ValueRange strides, @@ -1825,7 +1825,7 @@ private: } // Apply slice given the base address, extents and strides of the input box. - mlir::LogicalResult + llvm::LogicalResult sliceBox(fir::cg::XReboxOp rebox, mlir::Type destBoxTy, mlir::Value dest, mlir::Value base, mlir::ValueRange inputExtents, mlir::ValueRange inputStrides, mlir::ValueRange operands, @@ -1908,7 +1908,7 @@ private: /// Apply a new shape to the data described by a box given the base address, /// extents and strides of the box. - mlir::LogicalResult + llvm::LogicalResult reshapeBox(fir::cg::XReboxOp rebox, mlir::Type destBoxTy, mlir::Value dest, mlir::Value base, mlir::ValueRange inputExtents, mlir::ValueRange inputStrides, mlir::ValueRange operands, @@ -1960,7 +1960,7 @@ private: struct EmboxProcOpConversion : public fir::FIROpConversion<fir::EmboxProcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::EmboxProcOp emboxproc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { TODO(emboxproc.getLoc(), "fir.emboxproc codegen"); @@ -2027,7 +2027,7 @@ struct ExtractValueOpConversion public ValueOpCommon { using FIROpAndTypeConversion::FIROpAndTypeConversion; - mlir::LogicalResult + llvm::LogicalResult doRewrite(fir::ExtractValueOp extractVal, mlir::Type ty, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -2046,7 +2046,7 @@ struct InsertValueOpConversion public ValueOpCommon { using FIROpAndTypeConversion::FIROpAndTypeConversion; - mlir::LogicalResult + llvm::LogicalResult doRewrite(fir::InsertValueOp insertVal, mlir::Type ty, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -2074,7 +2074,7 @@ struct InsertOnRangeOpConversion } } - mlir::LogicalResult + llvm::LogicalResult doRewrite(fir::InsertOnRangeOp range, mlir::Type ty, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { @@ -2127,7 +2127,7 @@ struct XArrayCoorOpConversion : public fir::FIROpAndTypeConversion<fir::cg::XArrayCoorOp> { using FIROpAndTypeConversion::FIROpAndTypeConversion; - mlir::LogicalResult + llvm::LogicalResult doRewrite(fir::cg::XArrayCoorOp coor, mlir::Type llvmPtrTy, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { auto loc = coor.getLoc(); @@ -2296,7 +2296,7 @@ struct CoordinateOpConversion : public fir::FIROpAndTypeConversion<fir::CoordinateOp> { using FIROpAndTypeConversion::FIROpAndTypeConversion; - mlir::LogicalResult + llvm::LogicalResult doRewrite(fir::CoordinateOp coor, mlir::Type ty, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::ValueRange operands = adaptor.getOperands(); @@ -2396,7 +2396,7 @@ struct CoordinateOpConversion } private: - mlir::LogicalResult + llvm::LogicalResult doRewriteBox(fir::CoordinateOp coor, mlir::ValueRange operands, mlir::Location loc, mlir::ConversionPatternRewriter &rewriter) const { @@ -2482,7 +2482,7 @@ private: return mlir::success(); } - mlir::LogicalResult + llvm::LogicalResult doRewriteRefOrPtr(fir::CoordinateOp coor, mlir::Type llvmObjectTy, mlir::ValueRange operands, mlir::Location loc, mlir::ConversionPatternRewriter &rewriter) const { @@ -2593,7 +2593,7 @@ struct FieldIndexOpConversion : public fir::FIROpConversion<fir::FieldIndexOp> { using FIROpConversion::FIROpConversion; // NB: most field references should be resolved by this point - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::FieldIndexOp field, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { auto recTy = mlir::cast<fir::RecordType>(field.getOnType()); @@ -2633,7 +2633,7 @@ struct FieldIndexOpConversion : public fir::FIROpConversion<fir::FieldIndexOp> { struct FirEndOpConversion : public fir::FIROpConversion<fir::FirEndOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::FirEndOp firEnd, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { TODO(firEnd.getLoc(), "fir.end codegen"); @@ -2645,7 +2645,7 @@ struct FirEndOpConversion : public fir::FIROpConversion<fir::FirEndOp> { struct TypeDescOpConversion : public fir::FIROpConversion<fir::TypeDescOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::TypeDescOp typeDescOp, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Type inTy = typeDescOp.getInType(); @@ -2672,7 +2672,7 @@ struct TypeDescOpConversion : public fir::FIROpConversion<fir::TypeDescOp> { struct HasValueOpConversion : public fir::FIROpConversion<fir::HasValueOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::HasValueOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { rewriter.replaceOpWithNewOp<mlir::LLVM::ReturnOp>(op, @@ -2725,7 +2725,7 @@ static inline bool attributeTypeIsCompatible(mlir::MLIRContext *ctx, struct GlobalOpConversion : public fir::FIROpConversion<fir::GlobalOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::GlobalOp global, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { @@ -2867,7 +2867,7 @@ private: struct LoadOpConversion : public fir::FIROpConversion<fir::LoadOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::LoadOp load, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Type llvmLoadTy = convertObjectType(load.getType()); @@ -2925,7 +2925,7 @@ struct LoadOpConversion : public fir::FIROpConversion<fir::LoadOp> { struct NoReassocOpConversion : public fir::FIROpConversion<fir::NoReassocOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::NoReassocOp noreassoc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { rewriter.replaceOp(noreassoc, adaptor.getOperands()[0]); @@ -2986,7 +2986,7 @@ static void genCaseLadderStep(mlir::Location loc, mlir::Value cmp, struct SelectCaseOpConversion : public fir::FIROpConversion<fir::SelectCaseOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::SelectCaseOp caseOp, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { unsigned conds = caseOp.getNumConditions(); @@ -3101,7 +3101,7 @@ static void selectMatchAndRewrite(const fir::LLVMTypeConverter &lowering, struct SelectOpConversion : public fir::FIROpConversion<fir::SelectOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::SelectOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { selectMatchAndRewrite<fir::SelectOp>(lowerTy(), op, adaptor, rewriter); @@ -3113,7 +3113,7 @@ struct SelectOpConversion : public fir::FIROpConversion<fir::SelectOp> { struct SelectRankOpConversion : public fir::FIROpConversion<fir::SelectRankOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::SelectRankOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { selectMatchAndRewrite<fir::SelectRankOp>(lowerTy(), op, adaptor, rewriter); @@ -3125,7 +3125,7 @@ struct SelectRankOpConversion : public fir::FIROpConversion<fir::SelectRankOp> { struct SelectTypeOpConversion : public fir::FIROpConversion<fir::SelectTypeOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::SelectTypeOp select, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::emitError(select.getLoc(), @@ -3138,7 +3138,7 @@ struct SelectTypeOpConversion : public fir::FIROpConversion<fir::SelectTypeOp> { struct StoreOpConversion : public fir::FIROpConversion<fir::StoreOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::StoreOp store, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = store.getLoc(); @@ -3181,7 +3181,7 @@ namespace { struct UnboxCharOpConversion : public fir::FIROpConversion<fir::UnboxCharOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::UnboxCharOp unboxchar, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Type lenTy = convertType(unboxchar.getType(1)); @@ -3206,7 +3206,7 @@ struct UnboxCharOpConversion : public fir::FIROpConversion<fir::UnboxCharOp> { struct UnboxProcOpConversion : public fir::FIROpConversion<fir::UnboxProcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::UnboxProcOp unboxproc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { TODO(unboxproc.getLoc(), "fir.unboxproc codegen"); @@ -3218,7 +3218,7 @@ struct UnboxProcOpConversion : public fir::FIROpConversion<fir::UnboxProcOp> { struct UndefOpConversion : public fir::FIROpConversion<fir::UndefOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::UndefOp undef, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { rewriter.replaceOpWithNewOp<mlir::LLVM::UndefOp>( @@ -3230,7 +3230,7 @@ struct UndefOpConversion : public fir::FIROpConversion<fir::UndefOp> { struct ZeroOpConversion : public fir::FIROpConversion<fir::ZeroOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ZeroOp zero, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Type ty = convertType(zero.getType()); @@ -3244,7 +3244,7 @@ struct UnreachableOpConversion : public fir::FIROpConversion<fir::UnreachableOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::UnreachableOp unreach, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { rewriter.replaceOpWithNewOp<mlir::LLVM::UnreachableOp>(unreach); @@ -3261,7 +3261,7 @@ struct UnreachableOpConversion struct IsPresentOpConversion : public fir::FIROpConversion<fir::IsPresentOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::IsPresentOp isPresent, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Type idxTy = lowerTy().indexType(); @@ -3290,7 +3290,7 @@ struct IsPresentOpConversion : public fir::FIROpConversion<fir::IsPresentOp> { struct AbsentOpConversion : public fir::FIROpConversion<fir::AbsentOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AbsentOp absent, OpAdaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Type ty = convertType(absent.getType()); @@ -3350,7 +3350,7 @@ namespace { struct AddcOpConversion : public fir::FIROpConversion<fir::AddcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AddcOp addc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { // given: (x + iy) + (x' + iy') @@ -3365,7 +3365,7 @@ struct AddcOpConversion : public fir::FIROpConversion<fir::AddcOp> { struct SubcOpConversion : public fir::FIROpConversion<fir::SubcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::SubcOp subc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { // given: (x + iy) - (x' + iy') @@ -3381,7 +3381,7 @@ struct SubcOpConversion : public fir::FIROpConversion<fir::SubcOp> { struct MulcOpConversion : public fir::FIROpConversion<fir::MulcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::MulcOp mulc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { // TODO: Can we use a call to __muldc3 ? @@ -3415,7 +3415,7 @@ struct MulcOpConversion : public fir::FIROpConversion<fir::MulcOp> { struct DivcOpConversion : public fir::FIROpConversion<fir::DivcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::DivcOp divc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { // TODO: Can we use a call to __divdc3 instead? @@ -3455,7 +3455,7 @@ struct DivcOpConversion : public fir::FIROpConversion<fir::DivcOp> { struct NegcOpConversion : public fir::FIROpConversion<fir::NegcOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::NegcOp neg, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { // given: -(x + iy) @@ -3476,7 +3476,7 @@ struct NegcOpConversion : public fir::FIROpConversion<fir::NegcOp> { struct BoxOffsetOpConversion : public fir::FIROpConversion<fir::BoxOffsetOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::BoxOffsetOp boxOffset, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { @@ -3505,7 +3505,7 @@ struct MustBeDeadConversion : public fir::FIROpConversion<FromOp> { : fir::FIROpConversion<FromOp>(lowering, options) {} using OpAdaptor = typename FromOp::Adaptor; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(FromOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const final { if (!op->getUses().empty()) @@ -3519,7 +3519,7 @@ struct UnrealizedConversionCastOpConversion : public fir::FIROpConversion<mlir::UnrealizedConversionCastOp> { using FIROpConversion::FIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::UnrealizedConversionCastOp op, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { assert(op.getOutputs().getTypes().size() == 1 && "expect a single type"); @@ -3562,7 +3562,7 @@ class RenameMSVCLibmCallees public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::LLVM::CallOp op, mlir::PatternRewriter &rewriter) const override { rewriter.startOpModification(op); @@ -3581,7 +3581,7 @@ class RenameMSVCLibmFuncs public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::LLVM::LLVMFuncOp op, mlir::PatternRewriter &rewriter) const override { rewriter.startOpModification(op); diff --git a/flang/lib/Optimizer/CodeGen/CodeGenOpenMP.cpp b/flang/lib/Optimizer/CodeGen/CodeGenOpenMP.cpp index a6fa05f..48d6c000e 100644 --- a/flang/lib/Optimizer/CodeGen/CodeGenOpenMP.cpp +++ b/flang/lib/Optimizer/CodeGen/CodeGenOpenMP.cpp @@ -59,7 +59,7 @@ struct MapInfoOpConversion : public OpenMPFIROpConversion<mlir::omp::MapInfoOp> { using OpenMPFIROpConversion::OpenMPFIROpConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::omp::MapInfoOp curOp, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { const mlir::TypeConverter *converter = getTypeConverter(); diff --git a/flang/lib/Optimizer/CodeGen/PreCGRewrite.cpp b/flang/lib/Optimizer/CodeGen/PreCGRewrite.cpp index 6a0cd5e..c57f514 100644 --- a/flang/lib/Optimizer/CodeGen/PreCGRewrite.cpp +++ b/flang/lib/Optimizer/CodeGen/PreCGRewrite.cpp @@ -80,7 +80,7 @@ class EmboxConversion : public mlir::OpRewritePattern<fir::EmboxOp> { public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::EmboxOp embox, mlir::PatternRewriter &rewriter) const override { // If the embox does not include a shape, then do not convert it @@ -95,7 +95,7 @@ public: return mlir::failure(); } - mlir::LogicalResult rewriteStaticShape(fir::EmboxOp embox, + llvm::LogicalResult rewriteStaticShape(fir::EmboxOp embox, mlir::PatternRewriter &rewriter, fir::SequenceType seqTy) const { auto loc = embox.getLoc(); @@ -115,7 +115,7 @@ public: return mlir::success(); } - mlir::LogicalResult rewriteDynamicShape(fir::EmboxOp embox, + llvm::LogicalResult rewriteDynamicShape(fir::EmboxOp embox, mlir::PatternRewriter &rewriter, mlir::Value shapeVal) const { auto loc = embox.getLoc(); @@ -168,7 +168,7 @@ class ReboxConversion : public mlir::OpRewritePattern<fir::ReboxOp> { public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ReboxOp rebox, mlir::PatternRewriter &rewriter) const override { auto loc = rebox.getLoc(); @@ -227,7 +227,7 @@ class ArrayCoorConversion : public mlir::OpRewritePattern<fir::ArrayCoorOp> { public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ArrayCoorOp arrCoor, mlir::PatternRewriter &rewriter) const override { auto loc = arrCoor.getLoc(); @@ -277,7 +277,7 @@ public: DeclareOpConversion(mlir::MLIRContext *ctx, bool preserveDecl) : OpRewritePattern(ctx), preserveDeclare(preserveDecl) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::DeclareOp declareOp, mlir::PatternRewriter &rewriter) const override { if (!preserveDeclare) { @@ -316,7 +316,7 @@ class DummyScopeOpConversion public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::DummyScopeOp dummyScopeOp, mlir::PatternRewriter &rewriter) const override { rewriter.replaceOpWithNewOp<fir::UndefOp>(dummyScopeOp, diff --git a/flang/lib/Optimizer/CodeGen/TargetRewrite.cpp b/flang/lib/Optimizer/CodeGen/TargetRewrite.cpp index 8199c5e..561d700 100644 --- a/flang/lib/Optimizer/CodeGen/TargetRewrite.cpp +++ b/flang/lib/Optimizer/CodeGen/TargetRewrite.cpp @@ -667,7 +667,7 @@ public: /// Convert the type signatures on all the functions present in the module. /// As the type signature is being changed, this must also update the /// function itself to use any new arguments, etc. - mlir::LogicalResult convertTypes(mlir::ModuleOp mod) { + llvm::LogicalResult convertTypes(mlir::ModuleOp mod) { mlir::MLIRContext *ctx = mod->getContext(); auto targetCPU = specifics->getTargetCPU(); mlir::StringAttr targetCPUAttr = diff --git a/flang/lib/Optimizer/CodeGen/TypeConverter.cpp b/flang/lib/Optimizer/CodeGen/TypeConverter.cpp index 501a36f..ce86c62 100644 --- a/flang/lib/Optimizer/CodeGen/TypeConverter.cpp +++ b/flang/lib/Optimizer/CodeGen/TypeConverter.cpp @@ -170,7 +170,7 @@ mlir::Type LLVMTypeConverter::indexType() const { } // fir.type<name(p : TY'...){f : TY...}> --> llvm<"%name = { ty... }"> -std::optional<mlir::LogicalResult> LLVMTypeConverter::convertRecordType( +std::optional<llvm::LogicalResult> LLVMTypeConverter::convertRecordType( fir::RecordType derived, llvm::SmallVectorImpl<mlir::Type> &results) { auto name = fir::NameUniquer::dropTypeConversionMarkers(derived.getName()); auto st = mlir::LLVM::LLVMStructType::getIdentified(&getContext(), name); diff --git a/flang/lib/Optimizer/Dialect/CUF/CUFOps.cpp b/flang/lib/Optimizer/Dialect/CUF/CUFOps.cpp index 4fa1d39..53092be 100644 --- a/flang/lib/Optimizer/Dialect/CUF/CUFOps.cpp +++ b/flang/lib/Optimizer/Dialect/CUF/CUFOps.cpp @@ -51,7 +51,7 @@ void cuf::AllocOp::build(mlir::OpBuilder &builder, mlir::OperationState &result, } template <typename Op> -static mlir::LogicalResult checkCudaAttr(Op op) { +static llvm::LogicalResult checkCudaAttr(Op op) { if (op.getDataAttr() == cuf::DataAttribute::Device || op.getDataAttr() == cuf::DataAttribute::Managed || op.getDataAttr() == cuf::DataAttribute::Unified) @@ -59,19 +59,19 @@ static mlir::LogicalResult checkCudaAttr(Op op) { return op.emitOpError("expect device, managed or unified cuda attribute"); } -mlir::LogicalResult cuf::AllocOp::verify() { return checkCudaAttr(*this); } +llvm::LogicalResult cuf::AllocOp::verify() { return checkCudaAttr(*this); } //===----------------------------------------------------------------------===// // FreeOp //===----------------------------------------------------------------------===// -mlir::LogicalResult cuf::FreeOp::verify() { return checkCudaAttr(*this); } +llvm::LogicalResult cuf::FreeOp::verify() { return checkCudaAttr(*this); } //===----------------------------------------------------------------------===// // AllocateOp //===----------------------------------------------------------------------===// -mlir::LogicalResult cuf::AllocateOp::verify() { +llvm::LogicalResult cuf::AllocateOp::verify() { if (getPinned() && getStream()) return emitOpError("pinned and stream cannot appears at the same time"); if (!mlir::isa<fir::BaseBoxType>(fir::unwrapRefType(getBox().getType()))) @@ -94,7 +94,7 @@ mlir::LogicalResult cuf::AllocateOp::verify() { // DataTransferOp //===----------------------------------------------------------------------===// -mlir::LogicalResult cuf::DataTransferOp::verify() { +llvm::LogicalResult cuf::DataTransferOp::verify() { mlir::Type srcTy = getSrc().getType(); mlir::Type dstTy = getDst().getType(); if ((fir::isa_ref_type(srcTy) && fir::isa_ref_type(dstTy)) || @@ -114,7 +114,7 @@ mlir::LogicalResult cuf::DataTransferOp::verify() { // DeallocateOp //===----------------------------------------------------------------------===// -mlir::LogicalResult cuf::DeallocateOp::verify() { +llvm::LogicalResult cuf::DeallocateOp::verify() { if (!mlir::isa<fir::BaseBoxType>(fir::unwrapRefType(getBox().getType()))) return emitOpError( "expect box to be a reference to class or box type value"); @@ -225,7 +225,7 @@ void printCUFKernelLoopControl( p.printRegion(region, /*printEntryBlockArgs=*/false); } -mlir::LogicalResult cuf::KernelOp::verify() { +llvm::LogicalResult cuf::KernelOp::verify() { if (getLowerbound().size() != getUpperbound().size() || getLowerbound().size() != getStep().size()) return emitOpError( diff --git a/flang/lib/Optimizer/Dialect/FIROps.cpp b/flang/lib/Optimizer/Dialect/FIROps.cpp index 84711c5..a499a6e4 100644 --- a/flang/lib/Optimizer/Dialect/FIROps.cpp +++ b/flang/lib/Optimizer/Dialect/FIROps.cpp @@ -263,7 +263,7 @@ void fir::AllocaOp::print(mlir::OpAsmPrinter &p) { printAllocatableOp(p, *this); } -mlir::LogicalResult fir::AllocaOp::verify() { +llvm::LogicalResult fir::AllocaOp::verify() { llvm::SmallVector<llvm::StringRef> visited; if (verifyInType(getInType(), visited, numShapeOperands())) return emitOpError("invalid type for allocation"); @@ -339,7 +339,7 @@ void fir::AllocMemOp::print(mlir::OpAsmPrinter &p) { printAllocatableOp(p, *this); } -mlir::LogicalResult fir::AllocMemOp::verify() { +llvm::LogicalResult fir::AllocMemOp::verify() { llvm::SmallVector<llvm::StringRef> visited; if (verifyInType(getInType(), visited, numShapeOperands())) return emitOpError("invalid type for allocation"); @@ -375,7 +375,7 @@ static bool validTypeParams(mlir::Type dynTy, mlir::ValueRange typeParams) { return typeParams.size() == 0; } -mlir::LogicalResult fir::ArrayCoorOp::verify() { +llvm::LogicalResult fir::ArrayCoorOp::verify() { auto eleTy = fir::dyn_cast_ptrOrBoxEleTy(getMemref().getType()); auto arrTy = mlir::dyn_cast<fir::SequenceType>(eleTy); if (!arrTy) @@ -424,7 +424,7 @@ mlir::LogicalResult fir::ArrayCoorOp::verify() { // Pull in fir.embox and fir.rebox into fir.array_coor when possible. struct SimplifyArrayCoorOp : public mlir::OpRewritePattern<fir::ArrayCoorOp> { using mlir::OpRewritePattern<fir::ArrayCoorOp>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ArrayCoorOp op, mlir::PatternRewriter &rewriter) const override { mlir::Value memref = op.getMemref(); @@ -821,7 +821,7 @@ std::vector<mlir::Value> fir::ArrayLoadOp::getExtents() { return {}; } -mlir::LogicalResult fir::ArrayLoadOp::verify() { +llvm::LogicalResult fir::ArrayLoadOp::verify() { auto eleTy = fir::dyn_cast_ptrOrBoxEleTy(getMemref().getType()); auto arrTy = mlir::dyn_cast<fir::SequenceType>(eleTy); if (!arrTy) @@ -864,7 +864,7 @@ mlir::LogicalResult fir::ArrayLoadOp::verify() { // ArrayMergeStoreOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ArrayMergeStoreOp::verify() { +llvm::LogicalResult fir::ArrayMergeStoreOp::verify() { if (!mlir::isa<fir::ArrayLoadOp>(getOriginal().getDefiningOp())) return emitOpError("operand #0 must be result of a fir.array_load op"); if (auto sl = getSlice()) { @@ -914,7 +914,7 @@ mlir::Type validArraySubobject(A op) { return fir::applyPathToType(ty, op.getIndices()); } -mlir::LogicalResult fir::ArrayFetchOp::verify() { +llvm::LogicalResult fir::ArrayFetchOp::verify() { auto arrTy = mlir::cast<fir::SequenceType>(getSequence().getType()); auto indSize = getIndices().size(); if (indSize < arrTy.getDimension()) @@ -936,7 +936,7 @@ mlir::LogicalResult fir::ArrayFetchOp::verify() { // ArrayAccessOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ArrayAccessOp::verify() { +llvm::LogicalResult fir::ArrayAccessOp::verify() { auto arrTy = mlir::cast<fir::SequenceType>(getSequence().getType()); std::size_t indSize = getIndices().size(); if (indSize < arrTy.getDimension()) @@ -956,7 +956,7 @@ mlir::LogicalResult fir::ArrayAccessOp::verify() { // ArrayUpdateOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ArrayUpdateOp::verify() { +llvm::LogicalResult fir::ArrayUpdateOp::verify() { if (fir::isa_ref_type(getMerge().getType())) return emitOpError("does not support reference type for merge"); auto arrTy = mlir::cast<fir::SequenceType>(getSequence().getType()); @@ -978,7 +978,7 @@ mlir::LogicalResult fir::ArrayUpdateOp::verify() { // ArrayModifyOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ArrayModifyOp::verify() { +llvm::LogicalResult fir::ArrayModifyOp::verify() { auto arrTy = mlir::cast<fir::SequenceType>(getSequence().getType()); auto indSize = getIndices().size(); if (indSize < arrTy.getDimension()) @@ -1171,7 +1171,7 @@ void fir::CallOp::build(mlir::OpBuilder &builder, mlir::OperationState &result, // CharConvertOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::CharConvertOp::verify() { +llvm::LogicalResult fir::CharConvertOp::verify() { auto unwrap = [&](mlir::Type t) { t = fir::unwrapSequenceType(fir::dyn_cast_ptrEleTy(t)); return mlir::dyn_cast<fir::CharacterType>(t); @@ -1294,7 +1294,7 @@ void fir::ConstcOp::print(mlir::OpAsmPrinter &p) { p.printType(getType()); } -mlir::LogicalResult fir::ConstcOp::verify() { +llvm::LogicalResult fir::ConstcOp::verify() { if (!mlir::isa<fir::ComplexType>(getType())) return emitOpError("must be a !fir.complex type"); return mlir::success(); @@ -1427,7 +1427,7 @@ bool fir::ConvertOp::canBeConverted(mlir::Type inType, mlir::Type outType) { areVectorsCompatible(inType, outType); } -mlir::LogicalResult fir::ConvertOp::verify() { +llvm::LogicalResult fir::ConvertOp::verify() { if (canBeConverted(getValue().getType(), getType())) return mlir::success(); return emitOpError("invalid type conversion") @@ -1468,7 +1468,7 @@ mlir::ParseResult fir::CoordinateOp::parse(mlir::OpAsmParser &parser, return mlir::success(); } -mlir::LogicalResult fir::CoordinateOp::verify() { +llvm::LogicalResult fir::CoordinateOp::verify() { const mlir::Type refTy = getRef().getType(); if (fir::isa_ref_type(refTy)) { auto eleTy = fir::dyn_cast_ptrEleTy(refTy); @@ -1551,7 +1551,7 @@ mlir::LogicalResult fir::CoordinateOp::verify() { // DispatchOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::DispatchOp::verify() { +llvm::LogicalResult fir::DispatchOp::verify() { // Check that pass_arg_pos is in range of actual operands. pass_arg_pos is // unsigned so check for less than zero is not needed. if (getPassArgPos() && *getPassArgPos() > (getArgOperands().size() - 1)) @@ -1579,6 +1579,7 @@ void fir::TypeInfoOp::build(mlir::OpBuilder &builder, fir::RecordType parentType, llvm::ArrayRef<mlir::NamedAttribute> attrs) { result.addRegion(); + result.addRegion(); result.addAttribute(mlir::SymbolTable::getSymbolAttrName(), builder.getStringAttr(type.getName())); result.addAttribute(getTypeAttrName(result.name), mlir::TypeAttr::get(type)); @@ -1588,7 +1589,7 @@ void fir::TypeInfoOp::build(mlir::OpBuilder &builder, result.addAttributes(attrs); } -mlir::LogicalResult fir::TypeInfoOp::verify() { +llvm::LogicalResult fir::TypeInfoOp::verify() { if (!getDispatchTable().empty()) for (auto &op : getDispatchTable().front().without_terminator()) if (!mlir::isa<fir::DTEntryOp>(op)) @@ -1606,7 +1607,7 @@ mlir::LogicalResult fir::TypeInfoOp::verify() { // EmboxOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::EmboxOp::verify() { +llvm::LogicalResult fir::EmboxOp::verify() { auto eleTy = fir::dyn_cast_ptrEleTy(getMemref().getType()); bool isArray = false; if (auto seqTy = mlir::dyn_cast<fir::SequenceType>(eleTy)) { @@ -1642,7 +1643,7 @@ mlir::LogicalResult fir::EmboxOp::verify() { // EmboxCharOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::EmboxCharOp::verify() { +llvm::LogicalResult fir::EmboxCharOp::verify() { auto eleTy = fir::dyn_cast_ptrEleTy(getMemref().getType()); if (!mlir::dyn_cast_or_null<fir::CharacterType>(eleTy)) return mlir::failure(); @@ -1653,7 +1654,7 @@ mlir::LogicalResult fir::EmboxCharOp::verify() { // EmboxProcOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::EmboxProcOp::verify() { +llvm::LogicalResult fir::EmboxProcOp::verify() { // host bindings (optional) must be a reference to a tuple if (auto h = getHost()) { if (auto r = mlir::dyn_cast<fir::ReferenceType>(h.getType())) @@ -1691,7 +1692,7 @@ void fir::TypeDescOp::print(mlir::OpAsmPrinter &p) { p.printOptionalAttrDict(getOperation()->getAttrs(), {"in_type"}); } -mlir::LogicalResult fir::TypeDescOp::verify() { +llvm::LogicalResult fir::TypeDescOp::verify() { mlir::Type resultTy = getType(); if (auto tdesc = mlir::dyn_cast<fir::TypeDescType>(resultTy)) { if (tdesc.getOfTy() != getInType()) @@ -2036,7 +2037,7 @@ static void printCustomRangeSubscript(mlir::OpAsmPrinter &printer, } /// Range bounds must be nonnegative, and the range must not be empty. -mlir::LogicalResult fir::InsertOnRangeOp::verify() { +llvm::LogicalResult fir::InsertOnRangeOp::verify() { if (fir::hasDynamicSize(getSeq().getType())) return emitOpError("must have constant shape and size"); mlir::DenseIntElementsAttr coorAttr = getCoor(); @@ -2079,7 +2080,7 @@ struct UndoComplexPattern : public mlir::RewritePattern { UndoComplexPattern(mlir::MLIRContext *ctx) : mlir::RewritePattern("fir.insert_value", 2, ctx) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::Operation *op, mlir::PatternRewriter &rewriter) const override { auto insval = mlir::dyn_cast_or_null<fir::InsertValueOp>(op); @@ -2250,7 +2251,7 @@ mlir::ParseResult fir::IterWhileOp::parse(mlir::OpAsmParser &parser, return mlir::success(); } -mlir::LogicalResult fir::IterWhileOp::verify() { +llvm::LogicalResult fir::IterWhileOp::verify() { // Check that the body defines as single block argument for the induction // variable. auto *body = getBody(); @@ -2610,7 +2611,7 @@ fir::DoLoopOp fir::getForInductionVarOwner(mlir::Value val) { } // Lifted from loop.loop -mlir::LogicalResult fir::DoLoopOp::verify() { +llvm::LogicalResult fir::DoLoopOp::verify() { // Check that the body defines as single block argument for the induction // variable. auto *body = getBody(); @@ -2785,7 +2786,7 @@ static bool areCompatibleCharacterTypes(mlir::Type t1, mlir::Type t2) { return c1.getLen() == c2.getLen(); } -mlir::LogicalResult fir::ReboxOp::verify() { +llvm::LogicalResult fir::ReboxOp::verify() { auto inputBoxTy = getBox().getType(); if (fir::isa_unknown_size_box(inputBoxTy)) return emitOpError("box operand must not have unknown rank or type"); @@ -2883,7 +2884,7 @@ static bool areCompatibleAssumedRankElementType(mlir::Type inputEleTy, return false; } -mlir::LogicalResult fir::ReboxAssumedRankOp::verify() { +llvm::LogicalResult fir::ReboxAssumedRankOp::verify() { mlir::Type inputType = getBox().getType(); if (!mlir::isa<fir::BaseBoxType>(inputType) && !fir::isBoxAddress(inputType)) return emitOpError("input must be a box or box address"); @@ -2911,7 +2912,7 @@ void fir::ReboxAssumedRankOp::getEffects( // ResultOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ResultOp::verify() { +llvm::LogicalResult fir::ResultOp::verify() { auto *parentOp = (*this)->getParentOp(); auto results = parentOp->getResults(); auto operands = (*this)->getOperands(); @@ -2928,7 +2929,7 @@ mlir::LogicalResult fir::ResultOp::verify() { // SaveResultOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::SaveResultOp::verify() { +llvm::LogicalResult fir::SaveResultOp::verify() { auto resultType = getValue().getType(); if (resultType != fir::dyn_cast_ptrEleTy(getMemref().getType())) return emitOpError("value type must match memory reference type"); @@ -2992,7 +2993,7 @@ static constexpr llvm::StringRef getTargetOffsetAttr() { } template <typename OpT> -static mlir::LogicalResult verifyIntegralSwitchTerminator(OpT op) { +static llvm::LogicalResult verifyIntegralSwitchTerminator(OpT op) { if (!mlir::isa<mlir::IntegerType, mlir::IndexType, fir::IntegerType>( op.getSelector().getType())) return op.emitOpError("must be an integer"); @@ -3086,7 +3087,7 @@ static void printIntegralSwitchTerminator(OpT op, mlir::OpAsmPrinter &p) { // SelectOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::SelectOp::verify() { +llvm::LogicalResult fir::SelectOp::verify() { return verifyIntegralSwitchTerminator(*this); } @@ -3412,7 +3413,7 @@ void fir::SelectCaseOp::build(mlir::OpBuilder &builder, destOperands, attributes); } -mlir::LogicalResult fir::SelectCaseOp::verify() { +llvm::LogicalResult fir::SelectCaseOp::verify() { if (!mlir::isa<mlir::IntegerType, mlir::IndexType, fir::IntegerType, fir::LogicalType, fir::CharacterType>(getSelector().getType())) return emitOpError("must be an integer, character, or logical"); @@ -3443,7 +3444,7 @@ mlir::LogicalResult fir::SelectCaseOp::verify() { // SelectRankOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::SelectRankOp::verify() { +llvm::LogicalResult fir::SelectRankOp::verify() { return verifyIntegralSwitchTerminator(*this); } @@ -3608,7 +3609,7 @@ void fir::SelectTypeOp::print(mlir::OpAsmPrinter &p) { fir::SelectTypeOp::getOperandSegmentSizeAttr()}); } -mlir::LogicalResult fir::SelectTypeOp::verify() { +llvm::LogicalResult fir::SelectTypeOp::verify() { if (!mlir::isa<fir::BaseBoxType>(getSelector().getType())) return emitOpError("must be a fir.class or fir.box type"); if (auto boxType = mlir::dyn_cast<fir::BoxType>(getSelector().getType())) @@ -3670,7 +3671,7 @@ void fir::SelectTypeOp::build(mlir::OpBuilder &builder, // ShapeOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ShapeOp::verify() { +llvm::LogicalResult fir::ShapeOp::verify() { auto size = getExtents().size(); auto shapeTy = mlir::dyn_cast<fir::ShapeType>(getType()); assert(shapeTy && "must be a shape type"); @@ -3689,7 +3690,7 @@ void fir::ShapeOp::build(mlir::OpBuilder &builder, mlir::OperationState &result, // ShapeShiftOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ShapeShiftOp::verify() { +llvm::LogicalResult fir::ShapeShiftOp::verify() { auto size = getPairs().size(); if (size < 2 || size > 16 * 2) return emitOpError("incorrect number of args"); @@ -3706,7 +3707,7 @@ mlir::LogicalResult fir::ShapeShiftOp::verify() { // ShiftOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::ShiftOp::verify() { +llvm::LogicalResult fir::ShiftOp::verify() { auto size = getOrigins().size(); auto shiftTy = mlir::dyn_cast<fir::ShiftType>(getType()); assert(shiftTy && "must be a shift type"); @@ -3742,7 +3743,7 @@ unsigned fir::SliceOp::getOutputRank(mlir::ValueRange triples) { return rank; } -mlir::LogicalResult fir::SliceOp::verify() { +llvm::LogicalResult fir::SliceOp::verify() { auto size = getTriples().size(); if (size < 3 || size > 16 * 3) return emitOpError("incorrect number of args for triple"); @@ -3788,7 +3789,7 @@ void fir::StoreOp::print(mlir::OpAsmPrinter &p) { p << " : " << getMemref().getType(); } -mlir::LogicalResult fir::StoreOp::verify() { +llvm::LogicalResult fir::StoreOp::verify() { if (getValue().getType() != fir::dyn_cast_ptrEleTy(getMemref().getType())) return emitOpError("store value type must match memory reference type"); return mlir::success(); @@ -3920,7 +3921,7 @@ void fir::StringLitOp::print(mlir::OpAsmPrinter &p) { p.printType(getType()); } -mlir::LogicalResult fir::StringLitOp::verify() { +llvm::LogicalResult fir::StringLitOp::verify() { if (mlir::cast<mlir::IntegerAttr>(getSize()).getValue().isNegative()) return emitOpError("size must be non-negative"); if (auto xl = getOperation()->getAttr(fir::StringLitOp::xlist())) { @@ -3941,7 +3942,7 @@ mlir::LogicalResult fir::StringLitOp::verify() { // UnboxProcOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::UnboxProcOp::verify() { +llvm::LogicalResult fir::UnboxProcOp::verify() { if (auto eleTy = fir::dyn_cast_ptrEleTy(getRefTuple().getType())) if (mlir::isa<mlir::TupleType>(eleTy)) return mlir::success(); @@ -4067,7 +4068,7 @@ mlir::ParseResult fir::IfOp::parse(mlir::OpAsmParser &parser, return mlir::success(); } -mlir::LogicalResult fir::IfOp::verify() { +llvm::LogicalResult fir::IfOp::verify() { if (getNumResults() != 0 && getElseRegion().empty()) return emitOpError("must have an else block if defining values"); @@ -4109,7 +4110,7 @@ void fir::IfOp::resultToSourceOps(llvm::SmallVectorImpl<mlir::Value> &results, // BoxOffsetOp //===----------------------------------------------------------------------===// -mlir::LogicalResult fir::BoxOffsetOp::verify() { +llvm::LogicalResult fir::BoxOffsetOp::verify() { auto boxType = mlir::dyn_cast_or_null<fir::BaseBoxType>( fir::dyn_cast_ptrEleTy(getBoxRef().getType())); if (!boxType) @@ -4402,7 +4403,7 @@ mlir::Type fir::applyPathToType(mlir::Type eleTy, mlir::ValueRange path) { return eleTy; } -mlir::LogicalResult fir::DeclareOp::verify() { +llvm::LogicalResult fir::DeclareOp::verify() { auto fortranVar = mlir::cast<fir::FortranVariableOpInterface>(this->getOperation()); return fortranVar.verifyDeclareLikeOpImpl(getMemref()); diff --git a/flang/lib/Optimizer/Dialect/FIRType.cpp b/flang/lib/Optimizer/Dialect/FIRType.cpp index b3f2ec8..dbccacf 100644 --- a/flang/lib/Optimizer/Dialect/FIRType.cpp +++ b/flang/lib/Optimizer/Dialect/FIRType.cpp @@ -683,7 +683,7 @@ void fir::BoxProcType::print(mlir::AsmPrinter &printer) const { printer << "<" << getEleTy() << '>'; } -mlir::LogicalResult +llvm::LogicalResult BoxProcType::verify(llvm::function_ref<mlir::InFlightDiagnostic()> emitError, mlir::Type eleTy) { if (mlir::isa<mlir::FunctionType>(eleTy)) @@ -704,7 +704,7 @@ static bool cannotBePointerOrHeapElementType(mlir::Type eleTy) { // BoxType //===----------------------------------------------------------------------===// -mlir::LogicalResult +llvm::LogicalResult fir::BoxType::verify(llvm::function_ref<mlir::InFlightDiagnostic()> emitError, mlir::Type eleTy) { if (mlir::isa<fir::BaseBoxType>(eleTy)) @@ -773,7 +773,7 @@ void fir::CharacterType::print(mlir::AsmPrinter &printer) const { // ClassType //===----------------------------------------------------------------------===// -mlir::LogicalResult +llvm::LogicalResult fir::ClassType::verify(llvm::function_ref<mlir::InFlightDiagnostic()> emitError, mlir::Type eleTy) { if (mlir::isa<fir::RecordType, fir::SequenceType, fir::HeapType, @@ -820,7 +820,7 @@ void fir::HeapType::print(mlir::AsmPrinter &printer) const { printer << "<" << getEleTy() << '>'; } -mlir::LogicalResult +llvm::LogicalResult fir::HeapType::verify(llvm::function_ref<mlir::InFlightDiagnostic()> emitError, mlir::Type eleTy) { if (cannotBePointerOrHeapElementType(eleTy)) @@ -868,7 +868,7 @@ void fir::PointerType::print(mlir::AsmPrinter &printer) const { printer << "<" << getEleTy() << '>'; } -mlir::LogicalResult fir::PointerType::verify( +llvm::LogicalResult fir::PointerType::verify( llvm::function_ref<mlir::InFlightDiagnostic()> emitError, mlir::Type eleTy) { if (cannotBePointerOrHeapElementType(eleTy)) @@ -889,7 +889,7 @@ void fir::RealType::print(mlir::AsmPrinter &printer) const { printer << "<" << getFKind() << '>'; } -mlir::LogicalResult +llvm::LogicalResult fir::RealType::verify(llvm::function_ref<mlir::InFlightDiagnostic()> emitError, KindTy fKind) { // TODO @@ -1012,7 +1012,7 @@ detail::RecordTypeStorage const *fir::RecordType::uniqueKey() const { return getImpl(); } -mlir::LogicalResult fir::RecordType::verify( +llvm::LogicalResult fir::RecordType::verify( llvm::function_ref<mlir::InFlightDiagnostic()> emitError, llvm::StringRef name) { if (name.size() == 0) @@ -1047,7 +1047,7 @@ void fir::ReferenceType::print(mlir::AsmPrinter &printer) const { printer << "<" << getEleTy() << '>'; } -mlir::LogicalResult fir::ReferenceType::verify( +llvm::LogicalResult fir::ReferenceType::verify( llvm::function_ref<mlir::InFlightDiagnostic()> emitError, mlir::Type eleTy) { if (mlir::isa<ShapeType, ShapeShiftType, SliceType, FieldType, LenType, @@ -1121,7 +1121,7 @@ unsigned fir::SequenceType::getConstantRows() const { return count; } -mlir::LogicalResult fir::SequenceType::verify( +llvm::LogicalResult fir::SequenceType::verify( llvm::function_ref<mlir::InFlightDiagnostic()> emitError, llvm::ArrayRef<int64_t> shape, mlir::Type eleTy, mlir::AffineMapAttr layoutMap) { @@ -1196,7 +1196,7 @@ void fir::TypeDescType::print(mlir::AsmPrinter &printer) const { printer << "<" << getOfTy() << '>'; } -mlir::LogicalResult fir::TypeDescType::verify( +llvm::LogicalResult fir::TypeDescType::verify( llvm::function_ref<mlir::InFlightDiagnostic()> emitError, mlir::Type eleTy) { if (mlir::isa<BoxType, BoxCharType, BoxProcType, ShapeType, ShapeShiftType, @@ -1225,7 +1225,7 @@ void fir::VectorType::print(mlir::AsmPrinter &printer) const { printer << "<" << getLen() << ':' << getEleTy() << '>'; } -mlir::LogicalResult fir::VectorType::verify( +llvm::LogicalResult fir::VectorType::verify( llvm::function_ref<mlir::InFlightDiagnostic()> emitError, uint64_t len, mlir::Type eleTy) { if (!(fir::isa_real(eleTy) || fir::isa_integer(eleTy))) diff --git a/flang/lib/Optimizer/Dialect/FirAliasTagOpInterface.cpp b/flang/lib/Optimizer/Dialect/FirAliasTagOpInterface.cpp index 648f490..bf058c1 100644 --- a/flang/lib/Optimizer/Dialect/FirAliasTagOpInterface.cpp +++ b/flang/lib/Optimizer/Dialect/FirAliasTagOpInterface.cpp @@ -15,17 +15,17 @@ #include "flang/Optimizer/Dialect/FirAliasTagOpInterface.cpp.inc" -mlir::LogicalResult +llvm::LogicalResult fir::detail::verifyFirAliasTagOpInterface(mlir::Operation *op) { auto iface = mlir::cast<FirAliasTagOpInterface>(op); mlir::ArrayAttr tags = iface.getTBAATagsOrNull(); if (!tags) - return mlir::success(); + return llvm::success(); for (mlir::Attribute iter : tags) if (!mlir::isa<mlir::LLVM::TBAATagAttr>(iter)) return op->emitOpError("expected op to return array of ") << mlir::LLVM::TBAATagAttr::getMnemonic() << " attributes"; - return mlir::success(); + return llvm::success(); } diff --git a/flang/lib/Optimizer/Dialect/FortranVariableInterface.cpp b/flang/lib/Optimizer/Dialect/FortranVariableInterface.cpp index 70b1a2f..034f8c7 100644 --- a/flang/lib/Optimizer/Dialect/FortranVariableInterface.cpp +++ b/flang/lib/Optimizer/Dialect/FortranVariableInterface.cpp @@ -14,7 +14,7 @@ #include "flang/Optimizer/Dialect/FortranVariableInterface.cpp.inc" -mlir::LogicalResult +llvm::LogicalResult fir::FortranVariableOpInterface::verifyDeclareLikeOpImpl(mlir::Value memref) { const unsigned numExplicitTypeParams = getExplicitTypeParams().size(); mlir::Type memType = memref.getType(); diff --git a/flang/lib/Optimizer/Dialect/Support/KindMapping.cpp b/flang/lib/Optimizer/Dialect/Support/KindMapping.cpp index bcb1121..30c6030 100644 --- a/flang/lib/Optimizer/Dialect/Support/KindMapping.cpp +++ b/flang/lib/Optimizer/Dialect/Support/KindMapping.cpp @@ -177,7 +177,7 @@ static MatchResult parseInt(unsigned &result, const char *&ptr, return mlir::success(); } -static mlir::LogicalResult matchString(const char *&ptr, const char *endPtr, +static llvm::LogicalResult matchString(const char *&ptr, const char *endPtr, llvm::StringRef literal) { llvm::StringRef s(ptr, endPtr - ptr); if (s.starts_with(literal)) { @@ -351,7 +351,7 @@ std::string fir::KindMapping::mapToString() const { return result; } -mlir::LogicalResult +llvm::LogicalResult fir::KindMapping::setDefaultKinds(llvm::ArrayRef<KindTy> defs) { if (defs.empty()) { // generic front-end defaults diff --git a/flang/lib/Optimizer/HLFIR/IR/HLFIROps.cpp b/flang/lib/Optimizer/HLFIR/IR/HLFIROps.cpp index cbe789f..ae62aff 100644 --- a/flang/lib/Optimizer/HLFIR/IR/HLFIROps.cpp +++ b/flang/lib/Optimizer/HLFIR/IR/HLFIROps.cpp @@ -78,7 +78,7 @@ static bool isAllocatableBoxRef(mlir::Type type) { return boxType && mlir::isa<fir::HeapType>(boxType.getEleTy()); } -mlir::LogicalResult hlfir::AssignOp::verify() { +llvm::LogicalResult hlfir::AssignOp::verify() { mlir::Type lhsType = getLhs().getType(); if (isAllocatableAssignment() && !isAllocatableBoxRef(lhsType)) return emitOpError("lhs must be an allocatable when `realloc` is set"); @@ -137,7 +137,7 @@ void hlfir::DeclareOp::build(mlir::OpBuilder &builder, typeparams, dummy_scope, nameAttr, fortran_attrs, data_attr); } -mlir::LogicalResult hlfir::DeclareOp::verify() { +llvm::LogicalResult hlfir::DeclareOp::verify() { if (getMemref().getType() != getResult(1).getType()) return emitOpError("second result type must match input memref type"); mlir::Type hlfirVariableType = getHLFIRVariableType( @@ -278,7 +278,7 @@ static void printDesignatorComplexPart(mlir::OpAsmPrinter &p, } } -mlir::LogicalResult hlfir::DesignateOp::verify() { +llvm::LogicalResult hlfir::DesignateOp::verify() { mlir::Type memrefType = getMemref().getType(); mlir::Type baseType = getFortranElementOrSequenceType(memrefType); mlir::Type baseElementType = fir::unwrapSequenceType(baseType); @@ -433,7 +433,7 @@ mlir::LogicalResult hlfir::DesignateOp::verify() { // ParentComponentOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::ParentComponentOp::verify() { +llvm::LogicalResult hlfir::ParentComponentOp::verify() { mlir::Type baseType = hlfir::getFortranElementOrSequenceType(getMemref().getType()); auto maybeInputSeqType = mlir::dyn_cast<fir::SequenceType>(baseType); @@ -488,7 +488,7 @@ mlir::LogicalResult hlfir::ParentComponentOp::verify() { // LogicalReductionOp //===----------------------------------------------------------------------===// template <typename LogicalReductionOp> -static mlir::LogicalResult +static llvm::LogicalResult verifyLogicalReductionOp(LogicalReductionOp reductionOp) { mlir::Operation *op = reductionOp->getOperation(); @@ -539,7 +539,7 @@ verifyLogicalReductionOp(LogicalReductionOp reductionOp) { // AllOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::AllOp::verify() { +llvm::LogicalResult hlfir::AllOp::verify() { return verifyLogicalReductionOp<hlfir::AllOp *>(this); } @@ -554,7 +554,7 @@ void hlfir::AllOp::getEffects( // AnyOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::AnyOp::verify() { +llvm::LogicalResult hlfir::AnyOp::verify() { return verifyLogicalReductionOp<hlfir::AnyOp *>(this); } @@ -569,7 +569,7 @@ void hlfir::AnyOp::getEffects( // CountOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::CountOp::verify() { +llvm::LogicalResult hlfir::CountOp::verify() { mlir::Operation *op = getOperation(); auto results = op->getResultTypes(); @@ -626,7 +626,7 @@ getCharacterLengthIfStatic(mlir::Type t) { return std::nullopt; } -mlir::LogicalResult hlfir::ConcatOp::verify() { +llvm::LogicalResult hlfir::ConcatOp::verify() { if (getStrings().size() < 2) return emitOpError("must be provided at least two string operands"); unsigned kind = getCharacterKind(getResult().getType()); @@ -668,7 +668,7 @@ void hlfir::ConcatOp::getEffects( //===----------------------------------------------------------------------===// template <typename NumericalReductionOp> -static mlir::LogicalResult +static llvm::LogicalResult verifyArrayAndMaskForReductionOp(NumericalReductionOp reductionOp) { mlir::Value array = reductionOp->getArray(); mlir::Value mask = reductionOp->getMask(); @@ -707,7 +707,7 @@ verifyArrayAndMaskForReductionOp(NumericalReductionOp reductionOp) { } template <typename NumericalReductionOp> -static mlir::LogicalResult +static llvm::LogicalResult verifyNumericalReductionOp(NumericalReductionOp reductionOp) { mlir::Operation *op = reductionOp->getOperation(); auto results = op->getResultTypes(); @@ -760,7 +760,7 @@ verifyNumericalReductionOp(NumericalReductionOp reductionOp) { // ProductOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::ProductOp::verify() { +llvm::LogicalResult hlfir::ProductOp::verify() { return verifyNumericalReductionOp<hlfir::ProductOp *>(this); } @@ -776,7 +776,7 @@ void hlfir::ProductOp::getEffects( //===----------------------------------------------------------------------===// template <typename CharacterReductionOp> -static mlir::LogicalResult +static llvm::LogicalResult verifyCharacterReductionOp(CharacterReductionOp reductionOp) { mlir::Operation *op = reductionOp->getOperation(); auto results = op->getResultTypes(); @@ -821,7 +821,7 @@ verifyCharacterReductionOp(CharacterReductionOp reductionOp) { // MaxvalOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::MaxvalOp::verify() { +llvm::LogicalResult hlfir::MaxvalOp::verify() { mlir::Operation *op = getOperation(); auto results = op->getResultTypes(); @@ -845,7 +845,7 @@ void hlfir::MaxvalOp::getEffects( // MinvalOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::MinvalOp::verify() { +llvm::LogicalResult hlfir::MinvalOp::verify() { mlir::Operation *op = getOperation(); auto results = op->getResultTypes(); @@ -870,7 +870,7 @@ void hlfir::MinvalOp::getEffects( //===----------------------------------------------------------------------===// template <typename NumericalReductionOp> -static mlir::LogicalResult +static llvm::LogicalResult verifyResultForMinMaxLoc(NumericalReductionOp reductionOp) { mlir::Operation *op = reductionOp->getOperation(); auto results = op->getResultTypes(); @@ -908,7 +908,7 @@ verifyResultForMinMaxLoc(NumericalReductionOp reductionOp) { return mlir::success(); } -mlir::LogicalResult hlfir::MinlocOp::verify() { +llvm::LogicalResult hlfir::MinlocOp::verify() { auto res = verifyArrayAndMaskForReductionOp(this); if (failed(res)) return res; @@ -927,7 +927,7 @@ void hlfir::MinlocOp::getEffects( // MaxlocOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::MaxlocOp::verify() { +llvm::LogicalResult hlfir::MaxlocOp::verify() { auto res = verifyArrayAndMaskForReductionOp(this); if (failed(res)) return res; @@ -971,7 +971,7 @@ void hlfir::SetLengthOp::getEffects( // SumOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::SumOp::verify() { +llvm::LogicalResult hlfir::SumOp::verify() { return verifyNumericalReductionOp<hlfir::SumOp *>(this); } @@ -986,7 +986,7 @@ void hlfir::SumOp::getEffects( // DotProductOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::DotProductOp::verify() { +llvm::LogicalResult hlfir::DotProductOp::verify() { mlir::Value lhs = getLhs(); mlir::Value rhs = getRhs(); fir::SequenceType lhsTy = mlir::cast<fir::SequenceType>( @@ -1042,7 +1042,7 @@ void hlfir::DotProductOp::getEffects( // MatmulOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::MatmulOp::verify() { +llvm::LogicalResult hlfir::MatmulOp::verify() { mlir::Value lhs = getLhs(); mlir::Value rhs = getRhs(); fir::SequenceType lhsTy = mlir::cast<fir::SequenceType>( @@ -1111,7 +1111,7 @@ mlir::LogicalResult hlfir::MatmulOp::verify() { return mlir::success(); } -mlir::LogicalResult +llvm::LogicalResult hlfir::MatmulOp::canonicalize(MatmulOp matmulOp, mlir::PatternRewriter &rewriter) { // the only two uses of the transposed matrix should be for the hlfir.matmul @@ -1170,7 +1170,7 @@ void hlfir::MatmulOp::getEffects( // TransposeOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::TransposeOp::verify() { +llvm::LogicalResult hlfir::TransposeOp::verify() { mlir::Value array = getArray(); fir::SequenceType arrayTy = mlir::cast<fir::SequenceType>( hlfir::getFortranElementOrSequenceType(array.getType())); @@ -1212,7 +1212,7 @@ void hlfir::TransposeOp::getEffects( // MatmulTransposeOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::MatmulTransposeOp::verify() { +llvm::LogicalResult hlfir::MatmulTransposeOp::verify() { mlir::Value lhs = getLhs(); mlir::Value rhs = getRhs(); fir::SequenceType lhsTy = mlir::cast<fir::SequenceType>( @@ -1350,7 +1350,7 @@ void hlfir::EndAssociateOp::build(mlir::OpBuilder &builder, associate.getMustFreeStrorageFlag()); } -mlir::LogicalResult hlfir::EndAssociateOp::verify() { +llvm::LogicalResult hlfir::EndAssociateOp::verify() { mlir::Value var = getVar(); if (hlfir::mayHaveAllocatableComponent(var.getType()) && !hlfir::isFortranEntity(var)) @@ -1438,7 +1438,7 @@ mlir::Value hlfir::ElementalOp::getElementEntity() { return mlir::cast<hlfir::YieldElementOp>(getBody()->back()).getElementValue(); } -mlir::LogicalResult hlfir::ElementalOp::verify() { +llvm::LogicalResult hlfir::ElementalOp::verify() { mlir::Value mold = getMold(); hlfir::ExprType resultType = mlir::cast<hlfir::ExprType>(getType()); if (!!mold != resultType.isPolymorphic()) @@ -1476,7 +1476,7 @@ void hlfir::NullOp::build(mlir::OpBuilder &builder, // DestroyOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::DestroyOp::verify() { +llvm::LogicalResult hlfir::DestroyOp::verify() { if (mustFinalizeExpr()) { mlir::Value expr = getExpr(); hlfir::ExprType exprTy = mlir::cast<hlfir::ExprType>(expr.getType()); @@ -1517,7 +1517,7 @@ std::size_t hlfir::ShapeOfOp::getRank() { return shape.getRank(); } -mlir::LogicalResult hlfir::ShapeOfOp::verify() { +llvm::LogicalResult hlfir::ShapeOfOp::verify() { mlir::Value expr = getExpr(); hlfir::ExprType exprTy = mlir::cast<hlfir::ExprType>(expr.getType()); std::size_t exprRank = exprTy.getShape().size(); @@ -1532,7 +1532,7 @@ mlir::LogicalResult hlfir::ShapeOfOp::verify() { return mlir::success(); } -mlir::LogicalResult +llvm::LogicalResult hlfir::ShapeOfOp::canonicalize(ShapeOfOp shapeOf, mlir::PatternRewriter &rewriter) { // if extent information is available at compile time, immediately fold the @@ -1544,11 +1544,11 @@ hlfir::ShapeOfOp::canonicalize(ShapeOfOp shapeOf, mlir::Value shape = hlfir::genExprShape(rewriter, loc, expr); if (!shape) // shape information is not available at compile time - return mlir::LogicalResult::failure(); + return llvm::LogicalResult::failure(); rewriter.replaceAllUsesWith(shapeOf.getResult(), shape); rewriter.eraseOp(shapeOf); - return mlir::LogicalResult::success(); + return llvm::LogicalResult::success(); } //===----------------------------------------------------------------------===// @@ -1563,7 +1563,7 @@ void hlfir::GetExtentOp::build(mlir::OpBuilder &builder, build(builder, result, indexTy, shape, dimAttr); } -mlir::LogicalResult hlfir::GetExtentOp::verify() { +llvm::LogicalResult hlfir::GetExtentOp::verify() { fir::ShapeType shapeTy = mlir::cast<fir::ShapeType>(getShape().getType()); std::uint64_t rank = shapeTy.getRank(); llvm::APInt dim = getDim(); @@ -1640,7 +1640,7 @@ static mlir::Operation *getTerminator(mlir::Region ®ion) { return ®ion.back().back(); } -mlir::LogicalResult hlfir::RegionAssignOp::verify() { +llvm::LogicalResult hlfir::RegionAssignOp::verify() { if (!mlir::isa_and_nonnull<hlfir::YieldOp>(getTerminator(getRhsRegion()))) return emitOpError( "right-hand side region must be terminated by an hlfir.yield"); @@ -1692,7 +1692,7 @@ void hlfir::ElementalAddrOp::build(mlir::OpBuilder &builder, odsState.addRegion(); } -mlir::LogicalResult hlfir::ElementalAddrOp::verify() { +llvm::LogicalResult hlfir::ElementalAddrOp::verify() { hlfir::YieldOp yieldOp = mlir::dyn_cast_or_null<hlfir::YieldOp>(getTerminator(getBody())); if (!yieldOp) @@ -1729,7 +1729,7 @@ mlir::Region *hlfir::ElementalAddrOp::getElementCleanup() { // OrderedAssignmentTreeOpInterface //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::OrderedAssignmentTreeOpInterface::verifyImpl() { +llvm::LogicalResult hlfir::OrderedAssignmentTreeOpInterface::verifyImpl() { if (mlir::Region *body = getSubTreeRegion()) if (!body->empty()) for (mlir::Operation &op : body->front()) @@ -1812,7 +1812,7 @@ static bool yieldsLogical(mlir::Region ®ion, bool mustBeScalarI1) { hlfir::getFortranElementOrSequenceType(yieldType)); } -mlir::LogicalResult hlfir::ForallMaskOp::verify() { +llvm::LogicalResult hlfir::ForallMaskOp::verify() { if (!yieldsLogical(getMaskRegion(), /*mustBeScalarI1=*/true)) return emitOpError("mask region must yield a scalar i1"); mlir::Operation *op = getOperation(); @@ -1828,7 +1828,7 @@ mlir::LogicalResult hlfir::ForallMaskOp::verify() { //===----------------------------------------------------------------------===// template <typename ConcreteOp> -static mlir::LogicalResult verifyWhereAndElseWhereBody(ConcreteOp &concreteOp) { +static llvm::LogicalResult verifyWhereAndElseWhereBody(ConcreteOp &concreteOp) { for (mlir::Operation &op : concreteOp.getBody().front()) if (mlir::isa<hlfir::ForallOp>(op)) return concreteOp.emitOpError( @@ -1836,13 +1836,13 @@ static mlir::LogicalResult verifyWhereAndElseWhereBody(ConcreteOp &concreteOp) { return mlir::success(); } -mlir::LogicalResult hlfir::WhereOp::verify() { +llvm::LogicalResult hlfir::WhereOp::verify() { if (!yieldsLogical(getMaskRegion(), /*mustBeScalarI1=*/false)) return emitOpError("mask region must yield a logical array"); return verifyWhereAndElseWhereBody(*this); } -mlir::LogicalResult hlfir::ElseWhereOp::verify() { +llvm::LogicalResult hlfir::ElseWhereOp::verify() { if (!getMaskRegion().empty()) if (!yieldsLogical(getMaskRegion(), /*mustBeScalarI1=*/false)) return emitOpError( @@ -1854,7 +1854,7 @@ mlir::LogicalResult hlfir::ElseWhereOp::verify() { // ForallIndexOp //===----------------------------------------------------------------------===// -mlir::LogicalResult +llvm::LogicalResult hlfir::ForallIndexOp::canonicalize(hlfir::ForallIndexOp indexOp, mlir::PatternRewriter &rewriter) { for (mlir::Operation *user : indexOp->getResult(0).getUsers()) @@ -1878,7 +1878,7 @@ hlfir::ForallIndexOp::canonicalize(hlfir::ForallIndexOp indexOp, // CharExtremumOp //===----------------------------------------------------------------------===// -mlir::LogicalResult hlfir::CharExtremumOp::verify() { +llvm::LogicalResult hlfir::CharExtremumOp::verify() { if (getStrings().size() < 2) return emitOpError("must be provided at least two string operands"); unsigned kind = getCharacterKind(getResult().getType()); @@ -1922,7 +1922,7 @@ void hlfir::CharExtremumOp::getEffects( // GetLength //===----------------------------------------------------------------------===// -mlir::LogicalResult +llvm::LogicalResult hlfir::GetLengthOp::canonicalize(GetLengthOp getLength, mlir::PatternRewriter &rewriter) { mlir::Location loc = getLength.getLoc(); diff --git a/flang/lib/Optimizer/HLFIR/Transforms/BufferizeHLFIR.cpp b/flang/lib/Optimizer/HLFIR/Transforms/BufferizeHLFIR.cpp index e292b56..a70a6b3 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/BufferizeHLFIR.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/BufferizeHLFIR.cpp @@ -30,7 +30,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/Pass/Pass.h" #include "mlir/Pass/PassManager.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/ADT/TypeSwitch.h" @@ -180,7 +179,7 @@ struct AsExprOpConversion : public mlir::OpConversionPattern<hlfir::AsExprOp> { using mlir::OpConversionPattern<hlfir::AsExprOp>::OpConversionPattern; explicit AsExprOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::AsExprOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::AsExprOp asExpr, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = asExpr->getLoc(); @@ -205,7 +204,7 @@ struct ShapeOfOpConversion : public mlir::OpConversionPattern<hlfir::ShapeOfOp> { using mlir::OpConversionPattern<hlfir::ShapeOfOp>::OpConversionPattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ShapeOfOp shapeOf, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = shapeOf.getLoc(); @@ -237,7 +236,7 @@ struct ApplyOpConversion : public mlir::OpConversionPattern<hlfir::ApplyOp> { using mlir::OpConversionPattern<hlfir::ApplyOp>::OpConversionPattern; explicit ApplyOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::ApplyOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ApplyOp apply, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = apply->getLoc(); @@ -262,7 +261,7 @@ struct AssignOpConversion : public mlir::OpConversionPattern<hlfir::AssignOp> { using mlir::OpConversionPattern<hlfir::AssignOp>::OpConversionPattern; explicit AssignOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::AssignOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::AssignOp assign, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { llvm::SmallVector<mlir::Value> newOperands; @@ -279,7 +278,7 @@ struct ConcatOpConversion : public mlir::OpConversionPattern<hlfir::ConcatOp> { using mlir::OpConversionPattern<hlfir::ConcatOp>::OpConversionPattern; explicit ConcatOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::ConcatOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ConcatOp concat, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = concat->getLoc(); @@ -318,7 +317,7 @@ struct SetLengthOpConversion using mlir::OpConversionPattern<hlfir::SetLengthOp>::OpConversionPattern; explicit SetLengthOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::SetLengthOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::SetLengthOp setLength, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = setLength->getLoc(); @@ -351,7 +350,7 @@ struct GetLengthOpConversion using mlir::OpConversionPattern<hlfir::GetLengthOp>::OpConversionPattern; explicit GetLengthOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::GetLengthOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::GetLengthOp getLength, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = getLength->getLoc(); @@ -441,7 +440,7 @@ struct AssociateOpConversion using mlir::OpConversionPattern<hlfir::AssociateOp>::OpConversionPattern; explicit AssociateOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::AssociateOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::AssociateOp associate, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = associate->getLoc(); @@ -660,7 +659,7 @@ struct EndAssociateOpConversion using mlir::OpConversionPattern<hlfir::EndAssociateOp>::OpConversionPattern; explicit EndAssociateOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::EndAssociateOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::EndAssociateOp endAssociate, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = endAssociate->getLoc(); @@ -677,7 +676,7 @@ struct DestroyOpConversion using mlir::OpConversionPattern<hlfir::DestroyOp>::OpConversionPattern; explicit DestroyOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::DestroyOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::DestroyOp destroy, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { // If expr was bufferized on the heap, now is time to deallocate the buffer. @@ -706,7 +705,7 @@ struct NoReassocOpConversion using mlir::OpConversionPattern<hlfir::NoReassocOp>::OpConversionPattern; explicit NoReassocOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::NoReassocOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::NoReassocOp noreassoc, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = noreassoc->getLoc(); @@ -767,7 +766,7 @@ struct ElementalOpConversion // by the nesting level of ElementalOp's. setHasBoundedRewriteRecursion(); } - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ElementalOp elemental, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = elemental->getLoc(); @@ -860,7 +859,7 @@ struct CharExtremumOpConversion using mlir::OpConversionPattern<hlfir::CharExtremumOp>::OpConversionPattern; explicit CharExtremumOpConversion(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<hlfir::CharExtremumOp>{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::CharExtremumOp char_extremum, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = char_extremum->getLoc(); diff --git a/flang/lib/Optimizer/HLFIR/Transforms/ConvertToFIR.cpp b/flang/lib/Optimizer/HLFIR/Transforms/ConvertToFIR.cpp index 74bbab0..9820595 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/ConvertToFIR.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/ConvertToFIR.cpp @@ -38,7 +38,7 @@ class AssignOpConversion : public mlir::OpRewritePattern<hlfir::AssignOp> { public: explicit AssignOpConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::AssignOp assignOp, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = assignOp->getLoc(); @@ -231,7 +231,7 @@ public: return {res[0], res[1]}; } - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::CopyInOp copyInOp, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = copyInOp.getLoc(); @@ -249,7 +249,7 @@ public: explicit CopyOutOpConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::CopyOutOp copyOutOp, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = copyOutOp.getLoc(); @@ -290,7 +290,7 @@ public: explicit DeclareOpConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::DeclareOp declareOp, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = declareOp->getLoc(); @@ -428,7 +428,7 @@ public: explicit DesignateOpConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::DesignateOp designate, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = designate.getLoc(); @@ -648,7 +648,7 @@ public: explicit ParentComponentOpConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ParentComponentOp parentComponent, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = parentComponent.getLoc(); @@ -696,7 +696,7 @@ public: explicit NoReassocOpConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::NoReassocOp noreassoc, mlir::PatternRewriter &rewriter) const override { rewriter.replaceOpWithNewOp<fir::NoReassocOp>(noreassoc, @@ -709,7 +709,7 @@ class NullOpConversion : public mlir::OpRewritePattern<hlfir::NullOp> { public: explicit NullOpConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::NullOp nullop, mlir::PatternRewriter &rewriter) const override { rewriter.replaceOpWithNewOp<fir::ZeroOp>(nullop, nullop.getType()); @@ -722,7 +722,7 @@ class GetExtentOpConversion public: using mlir::OpRewritePattern<hlfir::GetExtentOp>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::GetExtentOp getExtentOp, mlir::PatternRewriter &rewriter) const override { mlir::Value shape = getExtentOp.getShape(); diff --git a/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp b/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp index 4b0b859..769e14b1 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp @@ -73,7 +73,7 @@ class InlineElementalConversion public: using mlir::OpRewritePattern<hlfir::ElementalOp>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ElementalOp elemental, mlir::PatternRewriter &rewriter) const override { std::optional<std::pair<hlfir::ApplyOp, hlfir::DestroyOp>> maybeTuple = diff --git a/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp b/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp index 0347bec..bfb4148 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp @@ -22,7 +22,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/Pass/Pass.h" #include "mlir/Pass/PassManager.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include <optional> @@ -235,7 +234,7 @@ protected: }; public: - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(OP operation, mlir::PatternRewriter &rewriter) const override { std::string opName; @@ -311,7 +310,7 @@ using AllOpConversion = HlfirReductionIntrinsicConversion<hlfir::AllOp>; struct CountOpConversion : public HlfirIntrinsicConversion<hlfir::CountOp> { using HlfirIntrinsicConversion<hlfir::CountOp>::HlfirIntrinsicConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::CountOp count, mlir::PatternRewriter &rewriter) const override { fir::FirOpBuilder builder{rewriter, count.getOperation()}; @@ -345,7 +344,7 @@ struct CountOpConversion : public HlfirIntrinsicConversion<hlfir::CountOp> { struct MatmulOpConversion : public HlfirIntrinsicConversion<hlfir::MatmulOp> { using HlfirIntrinsicConversion<hlfir::MatmulOp>::HlfirIntrinsicConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::MatmulOp matmul, mlir::PatternRewriter &rewriter) const override { fir::FirOpBuilder builder{rewriter, matmul.getOperation()}; @@ -376,7 +375,7 @@ struct DotProductOpConversion : public HlfirIntrinsicConversion<hlfir::DotProductOp> { using HlfirIntrinsicConversion<hlfir::DotProductOp>::HlfirIntrinsicConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::DotProductOp dotProduct, mlir::PatternRewriter &rewriter) const override { fir::FirOpBuilder builder{rewriter, dotProduct.getOperation()}; @@ -407,7 +406,7 @@ class TransposeOpConversion : public HlfirIntrinsicConversion<hlfir::TransposeOp> { using HlfirIntrinsicConversion<hlfir::TransposeOp>::HlfirIntrinsicConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::TransposeOp transpose, mlir::PatternRewriter &rewriter) const override { fir::FirOpBuilder builder{rewriter, transpose.getOperation()}; @@ -437,7 +436,7 @@ struct MatmulTransposeOpConversion using HlfirIntrinsicConversion< hlfir::MatmulTransposeOp>::HlfirIntrinsicConversion; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::MatmulTransposeOp multranspose, mlir::PatternRewriter &rewriter) const override { fir::FirOpBuilder builder{rewriter, multranspose.getOperation()}; diff --git a/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIROrderedAssignments.cpp b/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIROrderedAssignments.cpp index a1a89bb..85dd517 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIROrderedAssignments.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIROrderedAssignments.cpp @@ -1309,7 +1309,7 @@ static void lower(hlfir::OrderedAssignmentTreeOpInterface root, /// Shared rewrite entry point for all the ordered assignment tree root /// operations. It calls the scheduler and then apply the schedule. -static mlir::LogicalResult rewrite(hlfir::OrderedAssignmentTreeOpInterface root, +static llvm::LogicalResult rewrite(hlfir::OrderedAssignmentTreeOpInterface root, bool tryFusingAssignments, mlir::PatternRewriter &rewriter) { hlfir::Schedule schedule = @@ -1337,7 +1337,7 @@ public: explicit ForallOpConversion(mlir::MLIRContext *ctx, bool tryFusingAssignments) : OpRewritePattern{ctx}, tryFusingAssignments{tryFusingAssignments} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ForallOp forallOp, mlir::PatternRewriter &rewriter) const override { auto root = mlir::cast<hlfir::OrderedAssignmentTreeOpInterface>( @@ -1354,7 +1354,7 @@ public: explicit WhereOpConversion(mlir::MLIRContext *ctx, bool tryFusingAssignments) : OpRewritePattern{ctx}, tryFusingAssignments{tryFusingAssignments} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::WhereOp whereOp, mlir::PatternRewriter &rewriter) const override { auto root = mlir::cast<hlfir::OrderedAssignmentTreeOpInterface>( @@ -1370,7 +1370,7 @@ public: explicit RegionAssignConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::RegionAssignOp regionAssignOp, mlir::PatternRewriter &rewriter) const override { auto root = mlir::cast<hlfir::OrderedAssignmentTreeOpInterface>( diff --git a/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp b/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp index 1dfa2f9..c5b8095 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp @@ -89,7 +89,7 @@ private: public: using mlir::OpRewritePattern<hlfir::ElementalOp>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::ElementalOp elemental, mlir::PatternRewriter &rewriter) const override; }; @@ -465,7 +465,7 @@ ElementalAssignBufferization::findMatch(hlfir::ElementalOp elemental) { return match; } -mlir::LogicalResult ElementalAssignBufferization::matchAndRewrite( +llvm::LogicalResult ElementalAssignBufferization::matchAndRewrite( hlfir::ElementalOp elemental, mlir::PatternRewriter &rewriter) const { std::optional<MatchInfo> match = findMatch(elemental); if (!match) @@ -519,12 +519,12 @@ private: public: using mlir::OpRewritePattern<hlfir::AssignOp>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::AssignOp assign, mlir::PatternRewriter &rewriter) const override; }; -mlir::LogicalResult BroadcastAssignBufferization::matchAndRewrite( +llvm::LogicalResult BroadcastAssignBufferization::matchAndRewrite( hlfir::AssignOp assign, mlir::PatternRewriter &rewriter) const { // Since RHS is a scalar and LHS is an array, LHS must be allocated // in a conforming Fortran program, and LHS cannot be reallocated @@ -587,12 +587,12 @@ private: public: using mlir::OpRewritePattern<hlfir::AssignOp>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::AssignOp assign, mlir::PatternRewriter &rewriter) const override; }; -mlir::LogicalResult VariableAssignBufferization::matchAndRewrite( +llvm::LogicalResult VariableAssignBufferization::matchAndRewrite( hlfir::AssignOp assign, mlir::PatternRewriter &rewriter) const { if (assign.isAllocatableAssignment()) return rewriter.notifyMatchFailure(assign, "AssignOp may imply allocation"); @@ -716,7 +716,7 @@ class ReductionElementalConversion : public mlir::OpRewritePattern<Op> { public: using mlir::OpRewritePattern<Op>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(Op op, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = op.getLoc(); hlfir::ElementalOp elemental = @@ -817,7 +817,7 @@ class MinMaxlocElementalConversion : public mlir::OpRewritePattern<Op> { public: using mlir::OpRewritePattern<Op>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(Op mloc, mlir::PatternRewriter &rewriter) const override { if (!mloc.getMask() || mloc.getDim() || mloc.getBack()) return rewriter.notifyMatchFailure(mloc, diff --git a/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp b/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp index 6153c82..60b06437 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp @@ -35,7 +35,7 @@ class TransposeAsElementalConversion public: using mlir::OpRewritePattern<hlfir::TransposeOp>::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(hlfir::TransposeOp transpose, mlir::PatternRewriter &rewriter) const override { mlir::Location loc = transpose.getLoc(); diff --git a/flang/lib/Optimizer/Support/CMakeLists.txt b/flang/lib/Optimizer/Support/CMakeLists.txt index 55f5718..8794c24 100644 --- a/flang/lib/Optimizer/Support/CMakeLists.txt +++ b/flang/lib/Optimizer/Support/CMakeLists.txt @@ -5,6 +5,7 @@ add_flang_library(FIRSupport DataLayout.cpp InitFIR.cpp InternalNames.cpp + Utils.cpp DEPENDS FIROpsIncGen @@ -14,6 +15,7 @@ add_flang_library(FIRSupport ${extension_libs} LINK_LIBS + FIRDialect ${dialect_libs} ${extension_libs} MLIRBuiltinToLLVMIRTranslation diff --git a/flang/lib/Optimizer/Support/InternalNames.cpp b/flang/lib/Optimizer/Support/InternalNames.cpp index 65046ea..b2e2cd3 100644 --- a/flang/lib/Optimizer/Support/InternalNames.cpp +++ b/flang/lib/Optimizer/Support/InternalNames.cpp @@ -381,6 +381,15 @@ std::string fir::NameUniquer::getTypeDescriptorBindingTableName( return getDerivedTypeObjectName(mangledTypeName, bindingTableSeparator); } +std::string +fir::NameUniquer::getComponentInitName(llvm::StringRef mangledTypeName, + llvm::StringRef componentName) { + + std::string prefix = + getDerivedTypeObjectName(mangledTypeName, componentInitSeparator); + return prefix + "." + componentName.str(); +} + llvm::StringRef fir::NameUniquer::dropTypeConversionMarkers(llvm::StringRef mangledTypeName) { if (mangledTypeName.ends_with(boxprocSuffix)) diff --git a/flang/lib/Optimizer/Support/Utils.cpp b/flang/lib/Optimizer/Support/Utils.cpp new file mode 100644 index 0000000..5d663e2 --- /dev/null +++ b/flang/lib/Optimizer/Support/Utils.cpp @@ -0,0 +1,52 @@ +//===-- Utils.cpp ---------------------------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// +// +// Coding style: https://mlir.llvm.org/getting_started/DeveloperGuide/ +// +//===----------------------------------------------------------------------===// + +#include "flang/Optimizer/Support/Utils.h" +#include "flang/Optimizer/Dialect/FIROps.h" +#include "flang/Optimizer/Dialect/FIRType.h" +#include "flang/Optimizer/Support/InternalNames.h" + +fir::TypeInfoOp fir::lookupTypeInfoOp(fir::RecordType recordType, + mlir::ModuleOp module, + const mlir::SymbolTable *symbolTable) { + // fir.type_info was created with the mangled name of the derived type. + // It is the same as the name in the related fir.type, except when a pass + // lowered the fir.type (e.g., when lowering fir.boxproc type if the type has + // pointer procedure components), in which case suffix may have been added to + // the fir.type name. Get rid of them when looking up for the fir.type_info. + llvm::StringRef originalMangledTypeName = + fir::NameUniquer::dropTypeConversionMarkers(recordType.getName()); + return fir::lookupTypeInfoOp(originalMangledTypeName, module, symbolTable); +} + +fir::TypeInfoOp fir::lookupTypeInfoOp(llvm::StringRef name, + mlir::ModuleOp module, + const mlir::SymbolTable *symbolTable) { + if (symbolTable) + if (auto typeInfo = symbolTable->lookup<fir::TypeInfoOp>(name)) + return typeInfo; + return module.lookupSymbol<fir::TypeInfoOp>(name); +} + +std::optional<llvm::ArrayRef<int64_t>> fir::getComponentLowerBoundsIfNonDefault( + fir::RecordType recordType, llvm::StringRef component, + mlir::ModuleOp module, const mlir::SymbolTable *symbolTable) { + fir::TypeInfoOp typeInfo = + fir::lookupTypeInfoOp(recordType, module, symbolTable); + if (!typeInfo || typeInfo.getComponentInfo().empty()) + return std::nullopt; + for (auto componentInfo : + typeInfo.getComponentInfo().getOps<fir::DTComponentOp>()) + if (componentInfo.getName() == component) + return componentInfo.getLowerBounds(); + return std::nullopt; +} diff --git a/flang/lib/Optimizer/Transforms/AbstractResult.cpp b/flang/lib/Optimizer/Transforms/AbstractResult.cpp index 85472cd..3906aa5 100644 --- a/flang/lib/Optimizer/Transforms/AbstractResult.cpp +++ b/flang/lib/Optimizer/Transforms/AbstractResult.cpp @@ -84,7 +84,7 @@ public: CallConversion(mlir::MLIRContext *context, bool shouldBoxResult) : OpRewritePattern<Op>(context, 1), shouldBoxResult{shouldBoxResult} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(Op op, mlir::PatternRewriter &rewriter) const override { auto loc = op.getLoc(); auto result = op->getResult(0); @@ -192,7 +192,7 @@ public: using OpRewritePattern::OpRewritePattern; SaveResultOpConversion(mlir::MLIRContext *context) : OpRewritePattern(context) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::SaveResultOp op, mlir::PatternRewriter &rewriter) const override { rewriter.eraseOp(op); @@ -205,7 +205,7 @@ public: using OpRewritePattern::OpRewritePattern; ReturnOpConversion(mlir::MLIRContext *context, mlir::Value newArg) : OpRewritePattern(context), newArg{newArg} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::func::ReturnOp ret, mlir::PatternRewriter &rewriter) const override { auto loc = ret.getLoc(); @@ -258,7 +258,7 @@ public: using OpRewritePattern::OpRewritePattern; AddrOfOpConversion(mlir::MLIRContext *context, bool shouldBoxResult) : OpRewritePattern(context), shouldBoxResult{shouldBoxResult} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AddrOfOp addrOf, mlir::PatternRewriter &rewriter) const override { auto oldFuncTy = mlir::cast<mlir::FunctionType>(addrOf.getType()); @@ -432,4 +432,4 @@ public: }; } // end anonymous namespace -} // namespace fir
\ No newline at end of file +} // namespace fir diff --git a/flang/lib/Optimizer/Transforms/AffineDemotion.cpp b/flang/lib/Optimizer/Transforms/AffineDemotion.cpp index b4523a0..c416302d 100644 --- a/flang/lib/Optimizer/Transforms/AffineDemotion.cpp +++ b/flang/lib/Optimizer/Transforms/AffineDemotion.cpp @@ -95,7 +95,7 @@ public: class ConvertConversion : public mlir::OpRewritePattern<fir::ConvertOp> { public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ConvertOp op, mlir::PatternRewriter &rewriter) const override { if (mlir::isa<mlir::MemRefType>(op.getRes().getType())) { @@ -133,7 +133,7 @@ mlir::Type convertMemRef(mlir::MemRefType type) { class StdAllocConversion : public mlir::OpRewritePattern<memref::AllocOp> { public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(memref::AllocOp op, mlir::PatternRewriter &rewriter) const override { rewriter.replaceOpWithNewOp<fir::AllocaOp>(op, convertMemRef(op.getType()), diff --git a/flang/lib/Optimizer/Transforms/AffinePromotion.cpp b/flang/lib/Optimizer/Transforms/AffinePromotion.cpp index 7d0131a..43fccf5 100644 --- a/flang/lib/Optimizer/Transforms/AffinePromotion.cpp +++ b/flang/lib/Optimizer/Transforms/AffinePromotion.cpp @@ -453,7 +453,7 @@ public: AffineLoopConversion(mlir::MLIRContext *context, AffineFunctionAnalysis &afa) : OpRewritePattern(context), functionAnalysis(afa) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::DoLoopOp loop, mlir::PatternRewriter &rewriter) const override { LLVM_DEBUG(llvm::dbgs() << "AffineLoopConversion: rewriting loop:\n"; @@ -546,7 +546,7 @@ public: using OpRewritePattern::OpRewritePattern; AffineIfConversion(mlir::MLIRContext *context, AffineFunctionAnalysis &afa) : OpRewritePattern(context) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::IfOp op, mlir::PatternRewriter &rewriter) const override { LLVM_DEBUG(llvm::dbgs() << "AffineIfConversion: rewriting if:\n"; diff --git a/flang/lib/Optimizer/Transforms/ArrayValueCopy.cpp b/flang/lib/Optimizer/Transforms/ArrayValueCopy.cpp index ebc1862..8544d17 100644 --- a/flang/lib/Optimizer/Transforms/ArrayValueCopy.cpp +++ b/flang/lib/Optimizer/Transforms/ArrayValueCopy.cpp @@ -796,7 +796,7 @@ class ArrayLoadConversion : public mlir::OpRewritePattern<ArrayLoadOp> { public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ArrayLoadOp load, mlir::PatternRewriter &rewriter) const override { LLVM_DEBUG(llvm::dbgs() << "replace load " << load << " with undef.\n"); @@ -810,7 +810,7 @@ class ArrayMergeStoreConversion public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ArrayMergeStoreOp store, mlir::PatternRewriter &rewriter) const override { LLVM_DEBUG(llvm::dbgs() << "marking store " << store << " as dead.\n"); @@ -1248,7 +1248,7 @@ public: const OperationUseMapT &m) : ArrayUpdateConversionBase{ctx, a, m} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ArrayUpdateOp update, mlir::PatternRewriter &rewriter) const override { auto loc = update.getLoc(); @@ -1276,7 +1276,7 @@ public: const OperationUseMapT &m) : ArrayUpdateConversionBase{ctx, a, m} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ArrayModifyOp modify, mlir::PatternRewriter &rewriter) const override { auto loc = modify.getLoc(); @@ -1298,7 +1298,7 @@ public: const OperationUseMapT &m) : OpRewritePattern{ctx}, useMap{m} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ArrayFetchOp fetch, mlir::PatternRewriter &rewriter) const override { auto *op = fetch.getOperation(); @@ -1329,7 +1329,7 @@ public: const OperationUseMapT &m) : ArrayUpdateConversionBase{ctx, a, m} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ArrayAccessOp access, mlir::PatternRewriter &rewriter) const override { auto *op = access.getOperation(); @@ -1362,7 +1362,7 @@ public: explicit ArrayAmendConversion(mlir::MLIRContext *ctx) : OpRewritePattern{ctx} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ArrayAmendOp amend, mlir::PatternRewriter &rewriter) const override { auto *op = amend.getOperation(); diff --git a/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp b/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp index 5dfc5009..2c9c73e 100644 --- a/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp +++ b/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp @@ -63,7 +63,7 @@ public: : mlir::OpRewritePattern<fir::ReboxAssumedRankOp>(context), symbolTable{symbolTable}, kindMap{kindMap} {}; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::ReboxAssumedRankOp rebox, mlir::PatternRewriter &rewriter) const override { fir::FirOpBuilder builder{rewriter, kindMap, symbolTable}; @@ -123,7 +123,7 @@ public: : mlir::OpRewritePattern<fir::IsAssumedSizeOp>(context), symbolTable{symbolTable}, kindMap{kindMap} {}; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::IsAssumedSizeOp isAssumedSizeOp, mlir::PatternRewriter &rewriter) const override { fir::FirOpBuilder builder{rewriter, kindMap, symbolTable}; diff --git a/flang/lib/Optimizer/Transforms/CharacterConversion.cpp b/flang/lib/Optimizer/Transforms/CharacterConversion.cpp index 44baad7..aee7e8c 100644 --- a/flang/lib/Optimizer/Transforms/CharacterConversion.cpp +++ b/flang/lib/Optimizer/Transforms/CharacterConversion.cpp @@ -38,7 +38,7 @@ class CharacterConvertConversion public: using OpRewritePattern::OpRewritePattern; - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::CharConvertOp conv, mlir::PatternRewriter &rewriter) const override { auto kindMap = fir::getKindMapping(conv->getParentOfType<mlir::ModuleOp>()); diff --git a/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp b/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp index f7074a7..7d0b8b3 100644 --- a/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp +++ b/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp @@ -37,7 +37,7 @@ public: CallOpRewriter(mlir::MLIRContext *ctx, const mlir::DominanceInfo &_di) : OpRewritePattern(ctx), di(_di) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::CallOp callOp, mlir::PatternRewriter &rewriter) const override { LLVM_DEBUG(llvm::dbgs() << "Processing call op: " << callOp << "\n"); diff --git a/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp b/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp index 1af5a68..3b79d6d 100644 --- a/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp +++ b/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp @@ -47,7 +47,7 @@ public: : mlir::OpRewritePattern<fir::DoLoopOp>(ctx), forceLoopToExecuteOnce(forceLoopToExecuteOnce), setNSW(setNSW) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(DoLoopOp loop, mlir::PatternRewriter &rewriter) const override { auto loc = loop.getLoc(); @@ -162,7 +162,7 @@ public: CfgIfConv(mlir::MLIRContext *ctx, bool forceLoopToExecuteOnce, bool setNSW) : mlir::OpRewritePattern<fir::IfOp>(ctx) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(IfOp ifOp, mlir::PatternRewriter &rewriter) const override { auto loc = ifOp.getLoc(); @@ -228,7 +228,7 @@ public: bool setNSW) : mlir::OpRewritePattern<fir::IterWhileOp>(ctx), setNSW(setNSW) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::IterWhileOp whileOp, mlir::PatternRewriter &rewriter) const override { auto loc = whileOp.getLoc(); diff --git a/flang/lib/Optimizer/Transforms/MemoryAllocation.cpp b/flang/lib/Optimizer/Transforms/MemoryAllocation.cpp index ada67b4..03b1ae8 100644 --- a/flang/lib/Optimizer/Transforms/MemoryAllocation.cpp +++ b/flang/lib/Optimizer/Transforms/MemoryAllocation.cpp @@ -100,7 +100,7 @@ public: llvm::ArrayRef<mlir::Operation *> rets) : OpRewritePattern(ctx), returnOps(rets) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AllocaOp alloca, mlir::PatternRewriter &rewriter) const override { auto loc = alloca.getLoc(); diff --git a/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp b/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp index 76c12d2..57f19f2 100644 --- a/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp +++ b/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp @@ -50,7 +50,7 @@ public: SelectTypeConv(mlir::MLIRContext *ctx) : mlir::OpConversionPattern<fir::SelectTypeOp>(ctx) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::SelectTypeOp selectType, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override; @@ -60,7 +60,7 @@ private: mlir::Type ty, mlir::ModuleOp mod, mlir::PatternRewriter &rewriter) const; - mlir::LogicalResult genTypeLadderStep(mlir::Location loc, + llvm::LogicalResult genTypeLadderStep(mlir::Location loc, mlir::Value selector, mlir::Attribute attr, mlir::Block *dest, std::optional<mlir::ValueRange> destOps, @@ -81,7 +81,7 @@ struct DispatchOpConv : public OpConversionPattern<fir::DispatchOp> { : mlir::OpConversionPattern<fir::DispatchOp>(ctx), bindingTables(bindingTables) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::DispatchOp dispatch, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const override { mlir::Location loc = dispatch.getLoc(); @@ -135,7 +135,7 @@ struct DispatchOpConv : public OpConversionPattern<fir::DispatchOp> { // %18 = fir.field_index proc, !fir.type<_QM__fortran_type_infoTbinding> // %19 = fir.coordinate_of %17, %18 : (!fir.ref<!fir.type<_QM__fortran_type_infoTbinding>>, !fir.field) -> !fir.ref<!fir.type<_QM__fortran_builtinsT__builtin_c_funptr>> // %20 = fir.field_index __address, !fir.type<_QM__fortran_builtinsT__builtin_c_funptr> - // %21 = fir.coordinate_of %19, %20 : (!fir.ref<!fir.type<_QM__fortran_builtinsT__builtin_c_funptr>>, !fir.field) -> !fir.ref<i64> + // %21 = fir.coordinate_of %19, %20 : (!fir.ref<!fir.type<_QM__fortran_builtinsT__builtin_c_funptr>>, !fir.field) -> !fir.ref<i64> // %22 = fir.load %21 : !fir.ref<i64> // %23 = fir.convert %22 : (i64) -> (() -> ()) // fir.call %23() : () -> () @@ -217,7 +217,7 @@ private: class PolymorphicOpConversion : public fir::impl::PolymorphicOpConversionBase<PolymorphicOpConversion> { public: - mlir::LogicalResult initialize(mlir::MLIRContext *ctx) override { + llvm::LogicalResult initialize(mlir::MLIRContext *ctx) override { return mlir::success(); } @@ -250,7 +250,7 @@ public: }; } // namespace -mlir::LogicalResult SelectTypeConv::matchAndRewrite( +llvm::LogicalResult SelectTypeConv::matchAndRewrite( fir::SelectTypeOp selectType, OpAdaptor adaptor, mlir::ConversionPatternRewriter &rewriter) const { auto operands = adaptor.getOperands(); @@ -341,7 +341,7 @@ mlir::LogicalResult SelectTypeConv::matchAndRewrite( return mlir::success(); } -mlir::LogicalResult SelectTypeConv::genTypeLadderStep( +llvm::LogicalResult SelectTypeConv::genTypeLadderStep( mlir::Location loc, mlir::Value selector, mlir::Attribute attr, mlir::Block *dest, std::optional<mlir::ValueRange> destOps, mlir::ModuleOp mod, mlir::PatternRewriter &rewriter, diff --git a/flang/lib/Optimizer/Transforms/StackArrays.cpp b/flang/lib/Optimizer/Transforms/StackArrays.cpp index 7157b55..e8fa70eb 100644 --- a/flang/lib/Optimizer/Transforms/StackArrays.cpp +++ b/flang/lib/Optimizer/Transforms/StackArrays.cpp @@ -25,7 +25,6 @@ #include "mlir/IR/Value.h" #include "mlir/Interfaces/LoopLikeInterface.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include "mlir/Transforms/Passes.h" #include "llvm/ADT/DenseMap.h" @@ -180,7 +179,7 @@ public: private: llvm::DenseMap<mlir::Operation *, AllocMemMap> funcMaps; - mlir::LogicalResult analyseFunction(mlir::Operation *func); + llvm::LogicalResult analyseFunction(mlir::Operation *func); }; /// Converts a fir.allocmem to a fir.alloca @@ -191,7 +190,7 @@ public: const StackArraysAnalysisWrapper::AllocMemMap &candidateOps) : OpRewritePattern(ctx), candidateOps{candidateOps} {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(fir::AllocMemOp allocmem, mlir::PatternRewriter &rewriter) const override; @@ -415,7 +414,7 @@ void AllocationAnalysis::processOperation(mlir::Operation *op) { visitOperationImpl(op, *before, after); } -mlir::LogicalResult +llvm::LogicalResult StackArraysAnalysisWrapper::analyseFunction(mlir::Operation *func) { assert(mlir::isa<mlir::func::FuncOp>(func)); size_t nAllocs = 0; @@ -507,7 +506,7 @@ static mlir::Value convertAllocationType(mlir::PatternRewriter &rewriter, return conv; } -mlir::LogicalResult +llvm::LogicalResult AllocMemConversion::matchAndRewrite(fir::AllocMemOp allocmem, mlir::PatternRewriter &rewriter) const { auto oldInsertionPt = rewriter.saveInsertionPoint(); diff --git a/flang/lib/Semantics/resolve-directives.cpp b/flang/lib/Semantics/resolve-directives.cpp index 0bf63a0..16658d5 100644 --- a/flang/lib/Semantics/resolve-directives.cpp +++ b/flang/lib/Semantics/resolve-directives.cpp @@ -1708,26 +1708,46 @@ void OmpAttributeVisitor::ResolveSeqLoopIndexInParallelOrTaskConstruct( // Use of DO CONCURRENT inside OpenMP construct is unspecified behavior // till OpenMP-5.0 standard. // In above both cases we skip the privatization of iteration variables. +// [OpenMP 5.1] DO CONCURRENT indices are private bool OmpAttributeVisitor::Pre(const parser::DoConstruct &x) { - // TODO:[OpenMP 5.1] DO CONCURRENT indices are private - if (x.IsDoNormal()) { - if (!dirContext_.empty() && GetContext().withinConstruct) { + if (!dirContext_.empty() && GetContext().withinConstruct) { + llvm::SmallVector<const parser::Name *> ivs; + if (x.IsDoNormal()) { const parser::Name *iv{GetLoopIndex(x)}; - if (iv && iv->symbol) { - if (!iv->symbol->test(Symbol::Flag::OmpPreDetermined)) { - ResolveSeqLoopIndexInParallelOrTaskConstruct(*iv); - } else { - // TODO: conflict checks with explicitly determined DSA - } - ordCollapseLevel--; - if (ordCollapseLevel) { - if (const auto *details{iv->symbol->detailsIf<HostAssocDetails>()}) { - const Symbol *tpSymbol = &details->symbol(); - if (tpSymbol->test(Symbol::Flag::OmpThreadprivate)) { - context_.Say(iv->source, - "Loop iteration variable %s is not allowed in THREADPRIVATE."_err_en_US, - iv->ToString()); - } + if (iv && iv->symbol) + ivs.push_back(iv); + } else if (x.IsDoConcurrent()) { + const Fortran::parser::LoopControl *loopControl = &*x.GetLoopControl(); + const Fortran::parser::LoopControl::Concurrent &concurrent = + std::get<Fortran::parser::LoopControl::Concurrent>(loopControl->u); + const Fortran::parser::ConcurrentHeader &concurrentHeader = + std::get<Fortran::parser::ConcurrentHeader>(concurrent.t); + const std::list<Fortran::parser::ConcurrentControl> &controls = + std::get<std::list<Fortran::parser::ConcurrentControl>>( + concurrentHeader.t); + for (const auto &control : controls) { + const parser::Name *iv{&std::get<0>(control.t)}; + if (iv && iv->symbol) + ivs.push_back(iv); + } + } + ordCollapseLevel--; + for (auto iv : ivs) { + if (!iv->symbol->test(Symbol::Flag::OmpPreDetermined)) { + ResolveSeqLoopIndexInParallelOrTaskConstruct(*iv); + } else { + // TODO: conflict checks with explicitly determined DSA + } + if (ordCollapseLevel) { + if (const auto *details{iv->symbol->detailsIf<HostAssocDetails>()}) { + const Symbol *tpSymbol = &details->symbol(); + // TODO: DoConcurrent won't capture the following check because a new + // symbol is declared in ResolveIndexName(), which will not have the + // OmpThreadprivate flag. + if (tpSymbol->test(Symbol::Flag::OmpThreadprivate)) { + context_.Say(iv->source, + "Loop iteration variable %s is not allowed in THREADPRIVATE."_err_en_US, + iv->ToString()); } } } diff --git a/flang/test/Fir/fir-ops.fir b/flang/test/Fir/fir-ops.fir index ecc5c4ec..b8ae566 100644 --- a/flang/test/Fir/fir-ops.fir +++ b/flang/test/Fir/fir-ops.fir @@ -460,6 +460,13 @@ fir.type_info @dispatch_tbl : !fir.type<dispatch_tbl{i:i32}> dispatch_table { // CHECK-LABEL: fir.type_info @test_type_info noinit nodestroy nofinal extends !fir.type<parent{i:i32}> : !fir.type<test_type_info{i:i32,j:f32}> fir.type_info @test_type_info noinit nodestroy nofinal extends !fir.type<parent{i:i32}> : !fir.type<test_type_info{i:i32,j:f32}> +// CHECK-LABEL: fir.type_info @cpinfo : !fir.type<cpinfo{comp_i:!fir.array<10x20xi32>}> component_info { +// CHECK: fir.dt_component "component_info" lbs [2, 3] +// CHECK: } +fir.type_info @cpinfo : !fir.type<cpinfo{comp_i:!fir.array<10x20xi32>}> component_info { + fir.dt_component "component_info" lbs [2, 3] +} + // CHECK-LABEL: func @compare_complex( // CHECK-SAME: [[VAL_151:%.*]]: !fir.complex<16>, [[VAL_152:%.*]]: !fir.complex<16>) { func.func @compare_complex(%a : !fir.complex<16>, %b : !fir.complex<16>) { diff --git a/flang/test/Lower/HLFIR/procedure-pointer.f90 b/flang/test/Lower/HLFIR/procedure-pointer.f90 index ce20f19..69b215e 100644 --- a/flang/test/Lower/HLFIR/procedure-pointer.f90 +++ b/flang/test/Lower/HLFIR/procedure-pointer.f90 @@ -128,7 +128,7 @@ use m ! CHECK: fir.store %[[VAL_2]] to %[[VAL_0]] : !fir.ref<!fir.boxproc<(!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>>>> ! CHECK: %[[VAL_3:.*]]:2 = hlfir.declare %[[VAL_0]] {fortran_attrs = #fir.var_attrs<pointer>, uniq_name = "_QFsub4Ep2"} : (!fir.ref<!fir.boxproc<(!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>>>>) -> (!fir.ref<!fir.boxproc<(!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>>>>, !fir.ref<!fir.boxproc<(!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>>>>) ! CHECK: %[[VAL_4:.*]] = fir.address_of(@_QPchar_func) : (!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>> -! CHECK: %[[VAL_12:.*]] = arith.constant -1 : index +! CHECK: %[[VAL_12:.*]] = arith.constant 0 : index ! CHECK: %[[VAL_5:.*]] = fir.emboxproc %[[VAL_4]] : ((!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>>) -> !fir.boxproc<() -> ()> ! CHECK: %[[VAL_6:.*]] = fir.convert %[[VAL_12]] : (index) -> i64 ! CHECK: %[[VAL_7:.*]] = fir.undefined tuple<!fir.boxproc<() -> ()>, i64> @@ -375,7 +375,7 @@ end subroutine ! CHECK-LABEL: fir.global internal @_QFsub1Ep7 : !fir.boxproc<(!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>>> { ! CHECK: %[[VAL_0:.*]] = fir.address_of(@_QPchar_func) : (!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>> -! CHECK: %[[VAL_11:.*]] = arith.constant -1 : index +! CHECK: %[[VAL_11:.*]] = arith.constant 0 : index ! CHECK: %[[VAL_1:.*]] = fir.emboxproc %[[VAL_0]] : ((!fir.ref<i32>) -> !fir.box<!fir.ptr<!fir.char<1,?>>>) -> !fir.boxproc<() -> ()> ! CHECK: %[[VAL_2:.*]] = fir.convert %[[VAL_11]] : (index) -> i64 ! CHECK: %[[VAL_3:.*]] = fir.undefined tuple<!fir.boxproc<() -> ()>, i64> diff --git a/flang/test/Lower/HLFIR/type-info-components.f90 b/flang/test/Lower/HLFIR/type-info-components.f90 new file mode 100644 index 0000000..ee36f9c --- /dev/null +++ b/flang/test/Lower/HLFIR/type-info-components.f90 @@ -0,0 +1,55 @@ +! Test generation of fir.dt_component +! RUN: bbc -emit-hlfir %s -o - | FileCheck %s +subroutine test_1(x) + integer, save, target :: my_target + procedure() :: my_proc + type :: sometype + integer :: i(-1:8) = 42 + integer :: j(1:8) + integer, allocatable :: alloc(:) + integer, pointer :: p => my_target + integer, pointer :: p2 => NULL() + integer, pointer :: p3 + procedure(), pointer, nopass :: proc_p => my_proc + procedure(), pointer, nopass :: proc_p2 => NULL() + procedure(), pointer, nopass :: proc_p3 + end type + type(sometype) :: x +end subroutine +! CHECK-LABEL: fir.type_info @_QFtest_1Tsometype +! CHECK-SAME component_info { +! CHECK: fir.dt_component "i" lbs [-1] init @_QFtest_1E.di.sometype.i +! CHECK-NOT: fir.dt_component "j" +! CHECK: fir.dt_component "p" init @_QFtest_1E.di.sometype.p +! CHECK: fir.dt_component "p2" init @_QFtest_1E.di.sometype.p2 +! CHECK: fir.dt_component "proc_p" init @_QPmy_proc +! CHECK: } + +subroutine test_nesting(x) + type some_sub_type + integer :: i = 42 + end type + type sometype2 + type(some_sub_type) :: nested + end type + type(sometype2) :: x +end subroutine +! CHECK-LABEL: fir.type_info @_QFtest_nestingTsome_sub_type +! CHECK-SAME component_info { +! CHECK: fir.dt_component "i" init @_QFtest_nestingE.di.some_sub_type.i +! CHECK: } + +! CHECK: fir.type_info @_QFtest_nestingTsometype2 +! CHECK-NOT: fir.dt_component + + +subroutine data_like(x) + type sometype3 + integer :: i/42/ + end type + type(sometype3) :: x +end subroutine +! CHECK-LABEL: fir.type_info @_QFdata_likeTsometype3 +! CHECK-SAME component_info { +! CHECK: fir.dt_component "i" init @_QFdata_likeE.di.sometype3.i +! CHECK: } diff --git a/flang/test/Lower/dummy-procedure-character.f90 b/flang/test/Lower/dummy-procedure-character.f90 index 72d5485..9a2710f 100644 --- a/flang/test/Lower/dummy-procedure-character.f90 +++ b/flang/test/Lower/dummy-procedure-character.f90 @@ -82,7 +82,7 @@ subroutine cannot_compute_len_yet() end function end interface ! CHECK: %[[VAL_0:.*]] = fir.address_of(@_QPbar4) : (!fir.ref<!fir.char<1,?>>, index, !fir.ref<i32>) -> !fir.boxchar<1> -! CHECK: %[[VAL_1:.*]] = arith.constant -1 : index +! CHECK: %[[VAL_1:.*]] = arith.constant 0 : index ! CHECK: %[[VAL_2:.*]] = fir.emboxproc %[[VAL_0]] : ((!fir.ref<!fir.char<1,?>>, index, !fir.ref<i32>) -> !fir.boxchar<1>) -> !fir.boxproc<() -> ()> ! CHECK: %[[VAL_3:.*]] = fir.convert %[[VAL_1]] : (index) -> i64 ! CHECK: %[[VAL_4:.*]] = fir.undefined tuple<!fir.boxproc<() -> ()>, i64> @@ -97,7 +97,7 @@ subroutine cannot_compute_len_yet_2() character(*) :: bar5 external :: bar5 ! CHECK: %[[VAL_0:.*]] = fir.address_of(@_QPbar5) : (!fir.ref<!fir.char<1,?>>, index) -> !fir.boxchar<1> -! CHECK: %[[VAL_1:.*]] = arith.constant -1 : index +! CHECK: %[[VAL_1:.*]] = arith.constant 0 : index ! CHECK: %[[VAL_2:.*]] = fir.emboxproc %[[VAL_0]] : ((!fir.ref<!fir.char<1,?>>, index) -> !fir.boxchar<1>) -> !fir.boxproc<() -> ()> ! CHECK: %[[VAL_3:.*]] = fir.convert %[[VAL_1]] : (index) -> i64 ! CHECK: %[[VAL_4:.*]] = fir.undefined tuple<!fir.boxproc<() -> ()>, i64> diff --git a/flang/test/Semantics/OpenMP/doconcurrent01.f90 b/flang/test/Semantics/OpenMP/doconcurrent01.f90 new file mode 100644 index 0000000..7e3bdce8 --- /dev/null +++ b/flang/test/Semantics/OpenMP/doconcurrent01.f90 @@ -0,0 +1,17 @@ +! RUN: %python %S/../test_symbols.py %s %flang_fc1 -fopenmp + +! OpenMP 5.1.1 +! DO Concurrent indices are private + +!DEF: /private_iv (Subroutine)Subprogram +subroutine private_iv + !DEF: /private_iv/i ObjectEntity INTEGER(4) + integer i + !$omp parallel default(private) + !$omp single + !DEF: /private_iv/OtherConstruct1/i (OmpPrivate, OmpPreDetermined) HostAssoc INTEGER(4) + do concurrent(i=1:2) + end do + !$omp end single + !$omp end parallel +end subroutine diff --git a/flang/tools/bbc/bbc.cpp b/flang/tools/bbc/bbc.cpp index 3485c14..e5e41ad 100644 --- a/flang/tools/bbc/bbc.cpp +++ b/flang/tools/bbc/bbc.cpp @@ -269,7 +269,7 @@ createTargetMachine(llvm::StringRef targetTriple, std::string &error) { /// of the pass manager, allowing it to be invoked as soon as it's /// required without impacting the main pass pipeline that may be invoked /// more than once for verification. -static mlir::LogicalResult runOpenMPPasses(mlir::ModuleOp mlirModule) { +static llvm::LogicalResult runOpenMPPasses(mlir::ModuleOp mlirModule) { mlir::PassManager pm(mlirModule->getName(), mlir::OpPassManager::Nesting::Implicit); fir::createOpenMPFIRPassPipeline(pm, enableOpenMPDevice); @@ -285,7 +285,7 @@ static mlir::LogicalResult runOpenMPPasses(mlir::ModuleOp mlirModule) { // Translate Fortran input to FIR, a dialect of MLIR. //===----------------------------------------------------------------------===// -static mlir::LogicalResult convertFortranSourceToMLIR( +static llvm::LogicalResult convertFortranSourceToMLIR( std::string path, Fortran::parser::Options options, const ProgramName &programPrefix, Fortran::semantics::SemanticsContext &semanticsContext, diff --git a/flang/tools/tco/tco.cpp b/flang/tools/tco/tco.cpp index 399ea13..34ac0e1 100644 --- a/flang/tools/tco/tco.cpp +++ b/flang/tools/tco/tco.cpp @@ -74,7 +74,7 @@ static void printModule(mlir::ModuleOp mod, raw_ostream &output) { } // compile a .fir file -static mlir::LogicalResult +static llvm::LogicalResult compileFIR(const mlir::PassPipelineCLParser &passPipeline) { // check that there is a file to load ErrorOr<std::unique_ptr<MemoryBuffer>> fileOrErr = diff --git a/libc/config/linux/aarch64/entrypoints.txt b/libc/config/linux/aarch64/entrypoints.txt index 07fdbfd..ff35e8f 100644 --- a/libc/config/linux/aarch64/entrypoints.txt +++ b/libc/config/linux/aarch64/entrypoints.txt @@ -508,11 +508,13 @@ if(LIBC_TYPES_HAS_FLOAT16) libc.src.math.canonicalizef16 libc.src.math.ceilf16 libc.src.math.copysignf16 + libc.src.math.f16addf libc.src.math.f16div libc.src.math.f16divf libc.src.math.f16fmaf libc.src.math.f16sqrt libc.src.math.f16sqrtf + libc.src.math.f16subf libc.src.math.fabsf16 libc.src.math.fdimf16 libc.src.math.floorf16 diff --git a/libc/config/linux/x86_64/entrypoints.txt b/libc/config/linux/x86_64/entrypoints.txt index 89f9011..3eefa12 100644 --- a/libc/config/linux/x86_64/entrypoints.txt +++ b/libc/config/linux/x86_64/entrypoints.txt @@ -538,6 +538,7 @@ if(LIBC_TYPES_HAS_FLOAT16) libc.src.math.canonicalizef16 libc.src.math.ceilf16 libc.src.math.copysignf16 + libc.src.math.f16addf libc.src.math.f16div libc.src.math.f16divf libc.src.math.f16divl @@ -547,6 +548,7 @@ if(LIBC_TYPES_HAS_FLOAT16) libc.src.math.f16sqrt libc.src.math.f16sqrtf libc.src.math.f16sqrtl + libc.src.math.f16subf libc.src.math.fabsf16 libc.src.math.fdimf16 libc.src.math.floorf16 diff --git a/libc/docs/math/index.rst b/libc/docs/math/index.rst index ccafb1f..e4da3d4 100644 --- a/libc/docs/math/index.rst +++ b/libc/docs/math/index.rst @@ -124,10 +124,14 @@ Basic Operations +------------------+------------------+-----------------+------------------------+----------------------+------------------------+------------------------+----------------------------+ | dsub | N/A | N/A | | N/A | | 7.12.14.2 | F.10.11 | +------------------+------------------+-----------------+------------------------+----------------------+------------------------+------------------------+----------------------------+ +| f16add | |check|\* | | | N/A | | 7.12.14.1 | F.10.11 | ++------------------+------------------+-----------------+------------------------+----------------------+------------------------+------------------------+----------------------------+ | f16div | |check|\* | |check|\* | |check|\* | N/A | |check| | 7.12.14.4 | F.10.11 | +------------------+------------------+-----------------+------------------------+----------------------+------------------------+------------------------+----------------------------+ | f16fma | |check| | |check| | |check| | N/A | |check| | 7.12.14.5 | F.10.11 | +------------------+------------------+-----------------+------------------------+----------------------+------------------------+------------------------+----------------------------+ +| f16sub | |check|\* | | | N/A | | 7.12.14.2 | F.10.11 | ++------------------+------------------+-----------------+------------------------+----------------------+------------------------+------------------------+----------------------------+ | fabs | |check| | |check| | |check| | |check| | |check| | 7.12.7.3 | F.10.4.3 | +------------------+------------------+-----------------+------------------------+----------------------+------------------------+------------------------+----------------------------+ | fadd | N/A | | | N/A | | 7.12.14.1 | F.10.11 | diff --git a/libc/newhdrgen/yaml_to_classes.py b/libc/newhdrgen/yaml_to_classes.py index 7159dd9..9b52c9c 100644 --- a/libc/newhdrgen/yaml_to_classes.py +++ b/libc/newhdrgen/yaml_to_classes.py @@ -10,7 +10,6 @@ import yaml -import re import argparse from pathlib import Path @@ -54,10 +53,10 @@ def yaml_to_classes(yaml_data): standards = (function_data.get("standards", None),) header.add_function( Function( - standards, function_data["return_type"], function_data["name"], arguments, + standards, guard, attributes, ) @@ -103,7 +102,79 @@ def fill_public_api(header_str, h_def_content): return h_def_content.replace("%%public_api()", header_str, 1) -def main(yaml_file, h_def_file, output_dir): +def parse_function_details(details): + """ + Parse function details from a list of strings and return a Function object. + + Args: + details: A list containing function details + + Returns: + Function: An instance of Function initialized with the details. + """ + return_type, name, arguments, standards, guard, attributes = details + standards = standards.split(",") if standards != "null" else [] + arguments = [arg.strip() for arg in arguments.split(",")] + attributes = attributes.split(",") if attributes != "null" else [] + + return Function( + return_type=return_type, + name=name, + arguments=arguments, + standards=standards, + guard=guard if guard != "null" else None, + attributes=attributes if attributes else [], + ) + + +def add_function_to_yaml(yaml_file, function_details): + """ + Add a function to the YAML file. + + Args: + yaml_file: The path to the YAML file. + function_details: A list containing function details (return_type, name, arguments, standards, guard, attributes). + """ + new_function = parse_function_details(function_details) + + with open(yaml_file, "r") as f: + yaml_data = yaml.safe_load(f) + + if "functions" not in yaml_data: + yaml_data["functions"] = [] + + function_dict = { + "name": new_function.name, + "standards": new_function.standards, + "return_type": new_function.return_type, + "arguments": [{"type": arg} for arg in new_function.arguments], + } + + if new_function.guard: + function_dict["guard"] = new_function.guard + + if new_function.attributes: + function_dict["attributes"] = new_function.attributes + + yaml_data["functions"].append(function_dict) + + class IndentYamlListDumper(yaml.Dumper): + def increase_indent(self, flow=False, indentless=False): + return super(IndentYamlListDumper, self).increase_indent(flow, False) + + with open(yaml_file, "w") as f: + yaml.dump( + yaml_data, + f, + Dumper=IndentYamlListDumper, + default_flow_style=False, + sort_keys=False, + ) + + print(f"Added function {new_function.name} to {yaml_file}") + + +def main(yaml_file, h_def_file, output_dir, add_function=None): """ Main function to generate header files from YAML and .h.def templates. @@ -111,8 +182,12 @@ def main(yaml_file, h_def_file, output_dir): yaml_file: Path to the YAML file containing header specification. h_def_file: Path to the .h.def template file. output_dir: Directory to output the generated header file. + add_function: Details of the function to be added to the YAML file (if any). """ + if add_function: + add_function_to_yaml(yaml_file, add_function) + header = load_yaml_file(yaml_file) with open(h_def_file, "r") as f: @@ -143,6 +218,19 @@ if __name__ == "__main__": default=".", help="Directory to output the generated header file", ) + parser.add_argument( + "--add_function", + nargs=6, + metavar=( + "RETURN_TYPE", + "NAME", + "ARGUMENTS", + "STANDARDS", + "GUARD", + "ATTRIBUTES", + ), + help="Add a function to the YAML file", + ) args = parser.parse_args() - main(args.yaml_file, args.h_def_file, args.output_dir) + main(args.yaml_file, args.h_def_file, args.output_dir, args.add_function) diff --git a/libc/spec/llvm_libc_ext.td b/libc/spec/llvm_libc_ext.td index 74a90a4..ff7831f 100644 --- a/libc/spec/llvm_libc_ext.td +++ b/libc/spec/llvm_libc_ext.td @@ -57,6 +57,10 @@ def LLVMLibcExt : StandardSpec<"llvm_libc_ext"> { [], // Types [], // Enumerations [ + GuardedFunctionSpec<"f16addf", RetValSpec<Float16Type>, [ArgSpec<FloatType>, ArgSpec<FloatType>], "LIBC_TYPES_HAS_FLOAT16">, + + GuardedFunctionSpec<"f16subf", RetValSpec<Float16Type>, [ArgSpec<FloatType>, ArgSpec<FloatType>], "LIBC_TYPES_HAS_FLOAT16">, + GuardedFunctionSpec<"f16div", RetValSpec<Float16Type>, [ArgSpec<DoubleType>, ArgSpec<DoubleType>], "LIBC_TYPES_HAS_FLOAT16">, GuardedFunctionSpec<"f16divf", RetValSpec<Float16Type>, [ArgSpec<FloatType>, ArgSpec<FloatType>], "LIBC_TYPES_HAS_FLOAT16">, GuardedFunctionSpec<"f16divl", RetValSpec<Float16Type>, [ArgSpec<LongDoubleType>, ArgSpec<LongDoubleType>], "LIBC_TYPES_HAS_FLOAT16">, diff --git a/libc/src/__support/FPUtil/generic/CMakeLists.txt b/libc/src/__support/FPUtil/generic/CMakeLists.txt index fb49fd0..c73f687 100644 --- a/libc/src/__support/FPUtil/generic/CMakeLists.txt +++ b/libc/src/__support/FPUtil/generic/CMakeLists.txt @@ -50,6 +50,25 @@ add_header_library( ) add_header_library( + add_sub + HDRS + add_sub.h + DEPENDS + libc.hdr.errno_macros + libc.hdr.fenv_macros + libc.src.__support.CPP.algorithm + libc.src.__support.CPP.bit + libc.src.__support.CPP.type_traits + libc.src.__support.FPUtil.basic_operations + libc.src.__support.FPUtil.fenv_impl + libc.src.__support.FPUtil.fp_bits + libc.src.__support.FPUtil.dyadic_float + libc.src.__support.FPUtil.rounding_mode + libc.src.__support.macros.attributes + libc.src.__support.macros.optimization +) + +add_header_library( div HDRS div.h diff --git a/libc/src/__support/FPUtil/generic/add_sub.h b/libc/src/__support/FPUtil/generic/add_sub.h new file mode 100644 index 0000000..a522dea --- /dev/null +++ b/libc/src/__support/FPUtil/generic/add_sub.h @@ -0,0 +1,206 @@ +//===-- Add and subtract IEEE 754 floating-point numbers --------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_SRC___SUPPORT_FPUTIL_GENERIC_ADD_SUB_H +#define LLVM_LIBC_SRC___SUPPORT_FPUTIL_GENERIC_ADD_SUB_H + +#include "hdr/errno_macros.h" +#include "hdr/fenv_macros.h" +#include "src/__support/CPP/algorithm.h" +#include "src/__support/CPP/bit.h" +#include "src/__support/CPP/type_traits.h" +#include "src/__support/FPUtil/BasicOperations.h" +#include "src/__support/FPUtil/FEnvImpl.h" +#include "src/__support/FPUtil/FPBits.h" +#include "src/__support/FPUtil/dyadic_float.h" +#include "src/__support/FPUtil/rounding_mode.h" +#include "src/__support/macros/attributes.h" +#include "src/__support/macros/optimization.h" + +namespace LIBC_NAMESPACE::fputil::generic { + +template <bool IsSub, typename OutType, typename InType> +LIBC_INLINE cpp::enable_if_t<cpp::is_floating_point_v<OutType> && + cpp::is_floating_point_v<InType> && + sizeof(OutType) <= sizeof(InType), + OutType> +add_or_sub(InType x, InType y) { + using OutFPBits = FPBits<OutType>; + using OutStorageType = typename OutFPBits::StorageType; + using InFPBits = FPBits<InType>; + using InStorageType = typename InFPBits::StorageType; + + constexpr int GUARD_BITS_LEN = 3; + constexpr int RESULT_FRACTION_LEN = InFPBits::FRACTION_LEN + GUARD_BITS_LEN; + constexpr int RESULT_MANTISSA_LEN = RESULT_FRACTION_LEN + 1; + + using DyadicFloat = + DyadicFloat<cpp::bit_ceil(static_cast<size_t>(RESULT_MANTISSA_LEN))>; + + InFPBits x_bits(x); + InFPBits y_bits(y); + + bool is_effectively_add = (x_bits.sign() == y_bits.sign()) != IsSub; + + if (LIBC_UNLIKELY(x_bits.is_inf_or_nan() || y_bits.is_inf_or_nan() || + x_bits.is_zero() || y_bits.is_zero())) { + if (x_bits.is_nan() || y_bits.is_nan()) { + if (x_bits.is_signaling_nan() || y_bits.is_signaling_nan()) + raise_except_if_required(FE_INVALID); + + if (x_bits.is_quiet_nan()) { + InStorageType x_payload = static_cast<InStorageType>(getpayload(x)); + if ((x_payload & ~(OutFPBits::FRACTION_MASK >> 1)) == 0) + return OutFPBits::quiet_nan(x_bits.sign(), + static_cast<OutStorageType>(x_payload)) + .get_val(); + } + + if (y_bits.is_quiet_nan()) { + InStorageType y_payload = static_cast<InStorageType>(getpayload(y)); + if ((y_payload & ~(OutFPBits::FRACTION_MASK >> 1)) == 0) + return OutFPBits::quiet_nan(y_bits.sign(), + static_cast<OutStorageType>(y_payload)) + .get_val(); + } + + return OutFPBits::quiet_nan().get_val(); + } + + if (x_bits.is_inf()) { + if (y_bits.is_inf()) { + if (!is_effectively_add) { + raise_except_if_required(FE_INVALID); + return OutFPBits::quiet_nan().get_val(); + } + + return OutFPBits::inf(x_bits.sign()).get_val(); + } + + return OutFPBits::inf(x_bits.sign()).get_val(); + } + + if (y_bits.is_inf()) + return OutFPBits::inf(y_bits.sign()).get_val(); + + if (x_bits.is_zero()) { + if (y_bits.is_zero()) { + switch (quick_get_round()) { + case FE_DOWNWARD: + return OutFPBits::zero(Sign::NEG).get_val(); + default: + return OutFPBits::zero(Sign::POS).get_val(); + } + } + + // volatile prevents Clang from converting tmp to OutType and then + // immediately back to InType before negating it, resulting in double + // rounding. + volatile InType tmp = y; + if constexpr (IsSub) + tmp = -tmp; + return static_cast<OutType>(tmp); + } + + if (y_bits.is_zero()) { + volatile InType tmp = y; + if constexpr (IsSub) + tmp = -tmp; + return static_cast<OutType>(tmp); + } + } + + InType x_abs = x_bits.abs().get_val(); + InType y_abs = y_bits.abs().get_val(); + + if (x_abs == y_abs && !is_effectively_add) { + switch (quick_get_round()) { + case FE_DOWNWARD: + return OutFPBits::zero(Sign::NEG).get_val(); + default: + return OutFPBits::zero(Sign::POS).get_val(); + } + } + + Sign result_sign = Sign::POS; + + if (x_abs > y_abs) { + result_sign = x_bits.sign(); + } else if (x_abs < y_abs) { + if (is_effectively_add) + result_sign = y_bits.sign(); + else if (y_bits.is_pos()) + result_sign = Sign::NEG; + } else if (is_effectively_add) { + result_sign = x_bits.sign(); + } + + InFPBits max_bits(cpp::max(x_abs, y_abs)); + InFPBits min_bits(cpp::min(x_abs, y_abs)); + + InStorageType result_mant; + + if (max_bits.is_subnormal()) { + // min_bits must be subnormal too. + + if (is_effectively_add) + result_mant = max_bits.get_mantissa() + min_bits.get_mantissa(); + else + result_mant = max_bits.get_mantissa() - min_bits.get_mantissa(); + + result_mant <<= GUARD_BITS_LEN; + } else { + InStorageType max_mant = max_bits.get_explicit_mantissa() << GUARD_BITS_LEN; + InStorageType min_mant = min_bits.get_explicit_mantissa() << GUARD_BITS_LEN; + int alignment = + max_bits.get_biased_exponent() - min_bits.get_biased_exponent(); + + InStorageType aligned_min_mant = + min_mant >> cpp::min(alignment, RESULT_MANTISSA_LEN); + bool aligned_min_mant_sticky; + + if (alignment <= 3) + aligned_min_mant_sticky = false; + else if (alignment <= InFPBits::FRACTION_LEN + 3) + aligned_min_mant_sticky = + (min_mant << (InFPBits::STORAGE_LEN - alignment)) != 0; + else + aligned_min_mant_sticky = true; + + if (is_effectively_add) + result_mant = max_mant + (aligned_min_mant | aligned_min_mant_sticky); + else + result_mant = max_mant - (aligned_min_mant | aligned_min_mant_sticky); + } + + int result_exp = max_bits.get_exponent() - RESULT_FRACTION_LEN; + DyadicFloat result(result_sign, result_exp, result_mant); + return result.template as<OutType, /*ShouldSignalExceptions=*/true>(); +} + +template <typename OutType, typename InType> +LIBC_INLINE cpp::enable_if_t<cpp::is_floating_point_v<OutType> && + cpp::is_floating_point_v<InType> && + sizeof(OutType) <= sizeof(InType), + OutType> +add(InType x, InType y) { + return add_or_sub</*IsSub=*/false, OutType>(x, y); +} + +template <typename OutType, typename InType> +LIBC_INLINE cpp::enable_if_t<cpp::is_floating_point_v<OutType> && + cpp::is_floating_point_v<InType> && + sizeof(OutType) <= sizeof(InType), + OutType> +sub(InType x, InType y) { + return add_or_sub</*IsSub=*/true, OutType>(x, y); +} + +} // namespace LIBC_NAMESPACE::fputil::generic + +#endif // LLVM_LIBC_SRC___SUPPORT_FPUTIL_GENERIC_ADD_SUB_H diff --git a/libc/src/math/CMakeLists.txt b/libc/src/math/CMakeLists.txt index 607051a..5b20913 100644 --- a/libc/src/math/CMakeLists.txt +++ b/libc/src/math/CMakeLists.txt @@ -99,6 +99,8 @@ add_math_entrypoint_object(exp10f) add_math_entrypoint_object(expm1) add_math_entrypoint_object(expm1f) +add_math_entrypoint_object(f16addf) + add_math_entrypoint_object(f16div) add_math_entrypoint_object(f16divf) add_math_entrypoint_object(f16divl) @@ -114,6 +116,8 @@ add_math_entrypoint_object(f16sqrtf) add_math_entrypoint_object(f16sqrtl) add_math_entrypoint_object(f16sqrtf128) +add_math_entrypoint_object(f16subf) + add_math_entrypoint_object(fabs) add_math_entrypoint_object(fabsf) add_math_entrypoint_object(fabsl) diff --git a/libc/src/math/f16addf.h b/libc/src/math/f16addf.h new file mode 100644 index 0000000..31d0e78 --- /dev/null +++ b/libc/src/math/f16addf.h @@ -0,0 +1,20 @@ +//===-- Implementation header for f16addf -----------------------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_SRC_MATH_F16ADDF_H +#define LLVM_LIBC_SRC_MATH_F16ADDF_H + +#include "src/__support/macros/properties/types.h" + +namespace LIBC_NAMESPACE { + +float16 f16addf(float x, float y); + +} // namespace LIBC_NAMESPACE + +#endif // LLVM_LIBC_SRC_MATH_F16ADDF_H diff --git a/libc/src/math/f16subf.h b/libc/src/math/f16subf.h new file mode 100644 index 0000000..1d04a4c --- /dev/null +++ b/libc/src/math/f16subf.h @@ -0,0 +1,20 @@ +//===-- Implementation header for f16subf -----------------------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_SRC_MATH_F16SUBF_H +#define LLVM_LIBC_SRC_MATH_F16SUBF_H + +#include "src/__support/macros/properties/types.h" + +namespace LIBC_NAMESPACE { + +float16 f16subf(float x, float y); + +} // namespace LIBC_NAMESPACE + +#endif // LLVM_LIBC_SRC_MATH_F16SUBF_H diff --git a/libc/src/math/generic/CMakeLists.txt b/libc/src/math/generic/CMakeLists.txt index 395d4f8..d6ea8c5 100644 --- a/libc/src/math/generic/CMakeLists.txt +++ b/libc/src/math/generic/CMakeLists.txt @@ -3796,6 +3796,32 @@ add_entrypoint_object( ) add_entrypoint_object( + f16addf + SRCS + f16addf.cpp + HDRS + ../f16addf.h + DEPENDS + libc.src.__support.macros.properties.types + libc.src.__support.FPUtil.generic.add_sub + COMPILE_OPTIONS + -O3 +) + +add_entrypoint_object( + f16subf + SRCS + f16subf.cpp + HDRS + ../f16subf.h + DEPENDS + libc.src.__support.macros.properties.types + libc.src.__support.FPUtil.generic.add_sub + COMPILE_OPTIONS + -O3 +) + +add_entrypoint_object( f16div SRCS f16div.cpp diff --git a/libc/src/math/generic/f16addf.cpp b/libc/src/math/generic/f16addf.cpp new file mode 100644 index 0000000..f1761a1 --- /dev/null +++ b/libc/src/math/generic/f16addf.cpp @@ -0,0 +1,19 @@ +//===-- Implementation of f16addf function --------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "src/math/f16addf.h" +#include "src/__support/FPUtil/generic/add_sub.h" +#include "src/__support/common.h" + +namespace LIBC_NAMESPACE { + +LLVM_LIBC_FUNCTION(float16, f16addf, (float x, float y)) { + return fputil::generic::add<float16>(x, y); +} + +} // namespace LIBC_NAMESPACE diff --git a/libc/src/math/generic/f16subf.cpp b/libc/src/math/generic/f16subf.cpp new file mode 100644 index 0000000..e4532a4 --- /dev/null +++ b/libc/src/math/generic/f16subf.cpp @@ -0,0 +1,19 @@ +//===-- Implementation of f16subf function --------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "src/math/f16subf.h" +#include "src/__support/FPUtil/generic/add_sub.h" +#include "src/__support/common.h" + +namespace LIBC_NAMESPACE { + +LLVM_LIBC_FUNCTION(float16, f16subf, (float x, float y)) { + return fputil::generic::sub<float16>(x, y); +} + +} // namespace LIBC_NAMESPACE diff --git a/libc/test/src/math/AddTest.h b/libc/test/src/math/AddTest.h new file mode 100644 index 0000000..df0ef66 --- /dev/null +++ b/libc/test/src/math/AddTest.h @@ -0,0 +1,74 @@ +//===-- Utility class to test different flavors of float add ----*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_TEST_SRC_MATH_ADDTEST_H +#define LLVM_LIBC_TEST_SRC_MATH_ADDTEST_H + +#include "test/UnitTest/FEnvSafeTest.h" +#include "test/UnitTest/FPMatcher.h" +#include "test/UnitTest/Test.h" +#include "utils/MPFRWrapper/MPFRUtils.h" + +namespace mpfr = LIBC_NAMESPACE::testing::mpfr; + +template <typename OutType, typename InType> +class AddTest : public LIBC_NAMESPACE::testing::FEnvSafeTest { + + struct InConstants { + DECLARE_SPECIAL_CONSTANTS(InType) + }; + + using InFPBits = typename InConstants::FPBits; + using InStorageType = typename InConstants::StorageType; + + static constexpr InStorageType IN_MAX_NORMAL_U = + InFPBits::max_normal().uintval(); + static constexpr InStorageType IN_MIN_NORMAL_U = + InFPBits::min_normal().uintval(); + static constexpr InStorageType IN_MAX_SUBNORMAL_U = + InFPBits::max_subnormal().uintval(); + static constexpr InStorageType IN_MIN_SUBNORMAL_U = + InFPBits::min_subnormal().uintval(); + +public: + typedef OutType (*AddFunc)(InType, InType); + + void test_subnormal_range(AddFunc func) { + constexpr InStorageType COUNT = 100'001; + constexpr InStorageType STEP = + (IN_MAX_SUBNORMAL_U - IN_MIN_SUBNORMAL_U) / COUNT; + for (InStorageType i = 0, v = 0, w = IN_MAX_SUBNORMAL_U; i <= COUNT; + ++i, v += STEP, w -= STEP) { + InType x = InFPBits(v).get_val(); + InType y = InFPBits(w).get_val(); + mpfr::BinaryInput<InType> input{x, y}; + EXPECT_MPFR_MATCH_ALL_ROUNDING(mpfr::Operation::Add, input, func(x, y), + 0.5); + } + } + + void test_normal_range(AddFunc func) { + constexpr InStorageType COUNT = 100'001; + constexpr InStorageType STEP = (IN_MAX_NORMAL_U - IN_MIN_NORMAL_U) / COUNT; + for (InStorageType i = 0, v = 0, w = IN_MAX_NORMAL_U; i <= COUNT; + ++i, v += STEP, w -= STEP) { + InType x = InFPBits(v).get_val(); + InType y = InFPBits(w).get_val(); + mpfr::BinaryInput<InType> input{x, y}; + EXPECT_MPFR_MATCH_ALL_ROUNDING(mpfr::Operation::Add, input, func(x, y), + 0.5); + } + } +}; + +#define LIST_ADD_TESTS(OutType, InType, func) \ + using LlvmLibcAddTest = AddTest<OutType, InType>; \ + TEST_F(LlvmLibcAddTest, SubnormalRange) { test_subnormal_range(&func); } \ + TEST_F(LlvmLibcAddTest, NormalRange) { test_normal_range(&func); } + +#endif // LLVM_LIBC_TEST_SRC_MATH_ADDTEST_H diff --git a/libc/test/src/math/CMakeLists.txt b/libc/test/src/math/CMakeLists.txt index c07c6d7..637e672 100644 --- a/libc/test/src/math/CMakeLists.txt +++ b/libc/test/src/math/CMakeLists.txt @@ -1824,6 +1824,18 @@ add_fp_unittest( ) add_fp_unittest( + fmul_test + NEED_MPFR + SUITE + libc-math-unittests + SRCS + fmul_test.cpp + HDRS + FMulTest.h + DEPENDS + libc.src.math.fmul +) +add_fp_unittest( asinhf_test NEED_MPFR SUITE @@ -1985,6 +1997,32 @@ add_fp_unittest( ) add_fp_unittest( + f16addf_test + NEED_MPFR + SUITE + libc-math-unittests + SRCS + f16addf_test.cpp + HDRS + AddTest.h + DEPENDS + libc.src.math.f16addf +) + +add_fp_unittest( + f16subf_test + NEED_MPFR + SUITE + libc-math-unittests + SRCS + f16subf_test.cpp + HDRS + SubTest.h + DEPENDS + libc.src.math.f16subf +) + +add_fp_unittest( f16div_test NEED_MPFR SUITE diff --git a/libc/test/src/math/FMulTest.h b/libc/test/src/math/FMulTest.h new file mode 100644 index 0000000..8ca33ea --- /dev/null +++ b/libc/test/src/math/FMulTest.h @@ -0,0 +1,121 @@ +//===-- Utility class to test fmul[f|l] -------------------------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_TEST_SRC_MATH_FMULTEST_H +#define LLVM_LIBC_TEST_SRC_MATH_FMULTEST_H + +#include "src/__support/FPUtil/FPBits.h" +#include "test/UnitTest/FEnvSafeTest.h" +#include "test/UnitTest/FPMatcher.h" +#include "test/UnitTest/Test.h" +#include "utils/MPFRWrapper/MPFRUtils.h" + +namespace mpfr = LIBC_NAMESPACE::testing::mpfr; + +template <typename OutType, typename InType> +class FmulMPFRTest : public LIBC_NAMESPACE::testing::FEnvSafeTest { + + DECLARE_SPECIAL_CONSTANTS(InType) + +public: + typedef OutType (*FMulFunc)(InType, InType); + + void testFMulMPFR(FMulFunc func) { + constexpr int N = 10; + mpfr::BinaryInput<InType> INPUTS[N] = { + {3.0, 5.0}, + {0x1.0p1, 0x1.0p-131}, + {0x1.0p2, 0x1.0p-129}, + {1.0, 1.0}, + {-0.0, -0.0}, + {-0.0, 0.0}, + {0.0, -0.0}, + {0x1.0p100, 0x1.0p100}, + {1.0, 1.0 + 0x1.0p-128 + 0x1.0p-149 + 0x1.0p-150}, + {1.0, 0x1.0p-128 + 0x1.0p-149 + 0x1.0p-150}}; + + for (int i = 0; i < N; ++i) { + InType x = INPUTS[i].x; + InType y = INPUTS[i].y; + ASSERT_MPFR_MATCH_ALL_ROUNDING(mpfr::Operation::Fmul, INPUTS[i], + func(x, y), 0.5); + } + } + + void testSpecialInputsMPFR(FMulFunc func) { + constexpr int N = 27; + mpfr::BinaryInput<InType> INPUTS[N] = {{inf, 0x1.0p-129}, + {0x1.0p-129, inf}, + {inf, 2.0}, + {3.0, inf}, + {0.0, 0.0}, + {neg_inf, aNaN}, + {aNaN, neg_inf}, + {neg_inf, neg_inf}, + {0.0, neg_inf}, + {neg_inf, 0.0}, + {neg_inf, 1.0}, + {1.0, neg_inf}, + {neg_inf, 0x1.0p-129}, + {0x1.0p-129, neg_inf}, + {0.0, 0x1.0p-129}, + {inf, 0.0}, + {0.0, inf}, + {0.0, aNaN}, + {2.0, aNaN}, + {0x1.0p-129, aNaN}, + {inf, aNaN}, + {aNaN, aNaN}, + {0.0, sNaN}, + {2.0, sNaN}, + {0x1.0p-129, sNaN}, + {inf, sNaN}, + {sNaN, sNaN}}; + + for (int i = 0; i < N; ++i) { + InType x = INPUTS[i].x; + InType y = INPUTS[i].y; + ASSERT_MPFR_MATCH_ALL_ROUNDING(mpfr::Operation::Fmul, INPUTS[i], + func(x, y), 0.5); + } + } + + void testNormalRange(FMulFunc func) { + using FPBits = LIBC_NAMESPACE::fputil::FPBits<InType>; + using StorageType = typename FPBits::StorageType; + static constexpr StorageType MAX_NORMAL = FPBits::max_normal().uintval(); + static constexpr StorageType MIN_NORMAL = FPBits::min_normal().uintval(); + + constexpr StorageType COUNT = 10'001; + constexpr StorageType STEP = (MAX_NORMAL - MIN_NORMAL) / COUNT; + for (int signs = 0; signs < 4; ++signs) { + for (StorageType v = MIN_NORMAL, w = MAX_NORMAL; + v <= MAX_NORMAL && w >= MIN_NORMAL; v += STEP, w -= STEP) { + InType x = FPBits(v).get_val(), y = FPBits(w).get_val(); + if (signs % 2 == 1) { + x = -x; + } + if (signs >= 2) { + y = -y; + } + + mpfr::BinaryInput<InType> input{x, y}; + ASSERT_MPFR_MATCH_ALL_ROUNDING(mpfr::Operation::Fmul, input, func(x, y), + 0.5); + } + } + } +}; + +#define LIST_FMUL_MPFR_TESTS(OutType, InType, func) \ + using LlvmLibcFmulTest = FmulMPFRTest<OutType, InType>; \ + TEST_F(LlvmLibcFmulTest, MulMpfr) { testFMulMPFR(&func); } \ + TEST_F(LlvmLibcFmulTest, NanInfMpfr) { testSpecialInputsMPFR(&func); } \ + TEST_F(LlvmLibcFmulTest, NormalRange) { testNormalRange(&func); } + +#endif // LLVM_LIBC_TEST_SRC_MATH_FMULTEST_H diff --git a/libc/test/src/math/SubTest.h b/libc/test/src/math/SubTest.h new file mode 100644 index 0000000..9b40353 --- /dev/null +++ b/libc/test/src/math/SubTest.h @@ -0,0 +1,74 @@ +//===-- Utility class to test different flavors of float sub ----*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_TEST_SRC_MATH_SUBTEST_H +#define LLVM_LIBC_TEST_SRC_MATH_SUBTEST_H + +#include "test/UnitTest/FEnvSafeTest.h" +#include "test/UnitTest/FPMatcher.h" +#include "test/UnitTest/Test.h" +#include "utils/MPFRWrapper/MPFRUtils.h" + +namespace mpfr = LIBC_NAMESPACE::testing::mpfr; + +template <typename OutType, typename InType> +class SubTest : public LIBC_NAMESPACE::testing::FEnvSafeTest { + + struct InConstants { + DECLARE_SPECIAL_CONSTANTS(InType) + }; + + using InFPBits = typename InConstants::FPBits; + using InStorageType = typename InConstants::StorageType; + + static constexpr InStorageType IN_MAX_NORMAL_U = + InFPBits::max_normal().uintval(); + static constexpr InStorageType IN_MIN_NORMAL_U = + InFPBits::min_normal().uintval(); + static constexpr InStorageType IN_MAX_SUBNORMAL_U = + InFPBits::max_subnormal().uintval(); + static constexpr InStorageType IN_MIN_SUBNORMAL_U = + InFPBits::min_subnormal().uintval(); + +public: + using SubFunc = OutType (*)(InType, InType); + + void test_subnormal_range(SubFunc func) { + constexpr InStorageType COUNT = 100'001; + constexpr InStorageType STEP = + (IN_MAX_SUBNORMAL_U - IN_MIN_SUBNORMAL_U) / COUNT; + for (InStorageType i = 0, v = 0, w = IN_MAX_SUBNORMAL_U; i <= COUNT; + ++i, v += STEP, w -= STEP) { + InType x = InFPBits(v).get_val(); + InType y = InFPBits(w).get_val(); + mpfr::BinaryInput<InType> input{x, y}; + EXPECT_MPFR_MATCH_ALL_ROUNDING(mpfr::Operation::Sub, input, func(x, y), + 0.5); + } + } + + void test_normal_range(SubFunc func) { + constexpr InStorageType COUNT = 100'001; + constexpr InStorageType STEP = (IN_MAX_NORMAL_U - IN_MIN_NORMAL_U) / COUNT; + for (InStorageType i = 0, v = 0, w = IN_MAX_NORMAL_U; i <= COUNT; + ++i, v += STEP, w -= STEP) { + InType x = InFPBits(v).get_val(); + InType y = InFPBits(w).get_val(); + mpfr::BinaryInput<InType> input{x, y}; + EXPECT_MPFR_MATCH_ALL_ROUNDING(mpfr::Operation::Sub, input, func(x, y), + 0.5); + } + } +}; + +#define LIST_SUB_TESTS(OutType, InType, func) \ + using LlvmLibcSubTest = SubTest<OutType, InType>; \ + TEST_F(LlvmLibcSubTest, SubnormalRange) { test_subnormal_range(&func); } \ + TEST_F(LlvmLibcSubTest, NormalRange) { test_normal_range(&func); } + +#endif // LLVM_LIBC_TEST_SRC_MATH_SUBTEST_H diff --git a/libc/test/src/math/f16addf_test.cpp b/libc/test/src/math/f16addf_test.cpp new file mode 100644 index 0000000..1e8b432 --- /dev/null +++ b/libc/test/src/math/f16addf_test.cpp @@ -0,0 +1,13 @@ +//===-- Unittests for f16addf ---------------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "AddTest.h" + +#include "src/math/f16addf.h" + +LIST_ADD_TESTS(float16, float, LIBC_NAMESPACE::f16addf) diff --git a/libc/test/src/math/f16subf_test.cpp b/libc/test/src/math/f16subf_test.cpp new file mode 100644 index 0000000..68ad948 --- /dev/null +++ b/libc/test/src/math/f16subf_test.cpp @@ -0,0 +1,13 @@ +//===-- Unittests for f16subf ---------------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "SubTest.h" + +#include "src/math/f16subf.h" + +LIST_SUB_TESTS(float16, float, LIBC_NAMESPACE::f16subf) diff --git a/libc/test/src/math/fmul_test.cpp b/libc/test/src/math/fmul_test.cpp new file mode 100644 index 0000000..16eaa1a --- /dev/null +++ b/libc/test/src/math/fmul_test.cpp @@ -0,0 +1,13 @@ +//===-- Unittests for fmul-------------------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===---------------------------------------------------------------------===// + +#include "FMulTest.h" + +#include "src/math/fmul.h" + +LIST_FMUL_MPFR_TESTS(float, double, LIBC_NAMESPACE::fmul) diff --git a/libc/test/src/math/smoke/AddTest.h b/libc/test/src/math/smoke/AddTest.h new file mode 100644 index 0000000..c713c5a --- /dev/null +++ b/libc/test/src/math/smoke/AddTest.h @@ -0,0 +1,156 @@ +//===-- Utility class to test different flavors of float add ----*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_TEST_SRC_MATH_SMOKE_ADDTEST_H +#define LLVM_LIBC_TEST_SRC_MATH_SMOKE_ADDTEST_H + +#include "hdr/errno_macros.h" +#include "hdr/fenv_macros.h" +#include "src/__support/FPUtil/BasicOperations.h" +#include "test/UnitTest/FEnvSafeTest.h" +#include "test/UnitTest/FPMatcher.h" +#include "test/UnitTest/Test.h" + +template <typename OutType, typename InType> +class AddTest : public LIBC_NAMESPACE::testing::FEnvSafeTest { + + DECLARE_SPECIAL_CONSTANTS(OutType) + + struct InConstants { + DECLARE_SPECIAL_CONSTANTS(InType) + }; + + using InFPBits = typename InConstants::FPBits; + using InStorageType = typename InConstants::StorageType; + + InConstants in; + +public: + using AddFunc = OutType (*)(InType, InType); + + void test_special_numbers(AddFunc func) { + EXPECT_FP_IS_NAN(func(aNaN, aNaN)); + EXPECT_FP_IS_NAN_WITH_EXCEPTION(func(sNaN, sNaN), FE_INVALID); + + InType qnan_42 = InFPBits::quiet_nan(Sign::POS, 0x42).get_val(); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(qnan_42, zero))); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(zero, qnan_42))); + + if constexpr (sizeof(OutType) < sizeof(InType)) { + InStorageType max_payload = InFPBits::FRACTION_MASK >> 1; + InType qnan_max = InFPBits::quiet_nan(Sign::POS, max_payload).get_val(); + EXPECT_FP_EQ(zero, + LIBC_NAMESPACE::fputil::getpayload(func(qnan_max, zero))); + EXPECT_FP_EQ(zero, + LIBC_NAMESPACE::fputil::getpayload(func(zero, qnan_max))); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(qnan_max, qnan_42))); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(qnan_42, qnan_max))); + } + + EXPECT_FP_EQ(inf, func(inf, zero)); + EXPECT_FP_EQ(neg_inf, func(neg_inf, zero)); + EXPECT_FP_EQ(inf, func(inf, neg_zero)); + EXPECT_FP_EQ(neg_inf, func(neg_inf, neg_zero)); + } + + void test_invalid_operations(AddFunc func) { + EXPECT_FP_IS_NAN_WITH_EXCEPTION(func(inf, neg_inf), FE_INVALID); + EXPECT_FP_IS_NAN_WITH_EXCEPTION(func(neg_inf, inf), FE_INVALID); + } + + void test_range_errors(AddFunc func) { + using namespace LIBC_NAMESPACE::fputil::testing; + + if (ForceRoundingMode r(RoundingMode::Nearest); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(inf, func(max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(-inf, func(neg_max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + + EXPECT_FP_EQ_WITH_EXCEPTION(zero, func(in.min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION( + neg_zero, func(in.neg_min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + + if (ForceRoundingMode r(RoundingMode::TowardZero); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(max_normal, func(max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_max_normal, + func(neg_max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + + EXPECT_FP_EQ_WITH_EXCEPTION(zero, func(in.min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION( + neg_zero, func(in.neg_min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + + if (ForceRoundingMode r(RoundingMode::Downward); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(max_normal, func(max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_FP_EQ_WITH_EXCEPTION(-inf, func(neg_max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + + EXPECT_FP_EQ_WITH_EXCEPTION(zero, func(in.min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION( + neg_min_denormal, func(in.neg_min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + + if (ForceRoundingMode r(RoundingMode::Upward); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(inf, func(max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_max_normal, + func(neg_max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + + EXPECT_FP_EQ_WITH_EXCEPTION(min_denormal, + func(in.min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION( + neg_zero, func(in.neg_min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + } + + void test_inexact_results(AddFunc func) { + func(InType(1.0), min_denormal); + EXPECT_FP_EXCEPTION(FE_INEXACT); + } +}; + +#define LIST_ADD_TESTS(OutType, InType, func) \ + using LlvmLibcAddTest = AddTest<OutType, InType>; \ + TEST_F(LlvmLibcAddTest, SpecialNumbers) { test_special_numbers(&func); } \ + TEST_F(LlvmLibcAddTest, InvalidOperations) { \ + test_invalid_operations(&func); \ + } \ + TEST_F(LlvmLibcAddTest, RangeErrors) { test_range_errors(&func); } \ + TEST_F(LlvmLibcAddTest, InexactResults) { test_inexact_results(&func); } + +#endif // LLVM_LIBC_TEST_SRC_MATH_SMOKE_ADDTEST_H diff --git a/libc/test/src/math/smoke/CMakeLists.txt b/libc/test/src/math/smoke/CMakeLists.txt index a363644..1b269ed 100644 --- a/libc/test/src/math/smoke/CMakeLists.txt +++ b/libc/test/src/math/smoke/CMakeLists.txt @@ -3644,6 +3644,35 @@ add_fp_unittest( ) add_fp_unittest( + f16addf_test + SUITE + libc-math-smoke-tests + SRCS + f16addf_test.cpp + HDRS + AddTest.h + DEPENDS + libc.hdr.errno_macros + libc.hdr.fenv_macros + libc.src.__support.FPUtil.basic_operations + libc.src.math.f16addf +) + +add_fp_unittest( + f16subf_test + SUITE + libc-math-smoke-tests + SRCS + f16subf_test.cpp + HDRS + SubTest.h + DEPENDS + libc.hdr.fenv_macros + libc.src.__support.FPUtil.basic_operations + libc.src.math.f16subf +) + +add_fp_unittest( f16div_test SUITE libc-math-smoke-tests diff --git a/libc/test/src/math/smoke/SubTest.h b/libc/test/src/math/smoke/SubTest.h new file mode 100644 index 0000000..a4b3822 --- /dev/null +++ b/libc/test/src/math/smoke/SubTest.h @@ -0,0 +1,158 @@ +//===-- Utility class to test different flavors of float sub ----*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIBC_TEST_SRC_MATH_SMOKE_SUBTEST_H +#define LLVM_LIBC_TEST_SRC_MATH_SMOKE_SUBTEST_H + +#include "hdr/fenv_macros.h" +#include "src/__support/FPUtil/BasicOperations.h" +#include "test/UnitTest/FEnvSafeTest.h" +#include "test/UnitTest/FPMatcher.h" +#include "test/UnitTest/Test.h" + +template <typename OutType, typename InType> +class SubTest : public LIBC_NAMESPACE::testing::FEnvSafeTest { + + DECLARE_SPECIAL_CONSTANTS(OutType) + + struct InConstants { + DECLARE_SPECIAL_CONSTANTS(InType) + }; + + using InFPBits = typename InConstants::FPBits; + using InStorageType = typename InConstants::StorageType; + + InConstants in; + +public: + using SubFunc = OutType (*)(InType, InType); + + void test_special_numbers(SubFunc func) { + EXPECT_FP_IS_NAN(func(aNaN, aNaN)); + EXPECT_FP_IS_NAN_WITH_EXCEPTION(func(sNaN, sNaN), FE_INVALID); + + InType qnan_42 = InFPBits::quiet_nan(Sign::POS, 0x42).get_val(); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(qnan_42, zero))); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(zero, qnan_42))); + + if constexpr (sizeof(OutType) < sizeof(InType)) { + InStorageType max_payload = InFPBits::FRACTION_MASK >> 1; + InType qnan_max = InFPBits::quiet_nan(Sign::POS, max_payload).get_val(); + EXPECT_FP_EQ(zero, + LIBC_NAMESPACE::fputil::getpayload(func(qnan_max, zero))); + EXPECT_FP_EQ(zero, + LIBC_NAMESPACE::fputil::getpayload(func(zero, qnan_max))); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(qnan_max, qnan_42))); + EXPECT_FP_EQ(InType(0x42.0p+0), + LIBC_NAMESPACE::fputil::getpayload(func(qnan_42, qnan_max))); + } + + EXPECT_FP_EQ(inf, func(inf, zero)); + EXPECT_FP_EQ(neg_inf, func(neg_inf, zero)); + EXPECT_FP_EQ(inf, func(inf, neg_zero)); + EXPECT_FP_EQ(neg_inf, func(neg_inf, neg_zero)); + } + + void test_invalid_operations(SubFunc func) { + EXPECT_FP_IS_NAN_WITH_EXCEPTION(func(inf, inf), FE_INVALID); + EXPECT_FP_IS_NAN_WITH_EXCEPTION(func(neg_inf, neg_inf), FE_INVALID); + } + + void test_range_errors(SubFunc func) { + using namespace LIBC_NAMESPACE::fputil::testing; + + if (ForceRoundingMode r(RoundingMode::Nearest); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(inf, func(max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(-inf, func(neg_max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + + EXPECT_FP_EQ_WITH_EXCEPTION(zero, + func(in.min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_zero, + func(in.neg_min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + + if (ForceRoundingMode r(RoundingMode::TowardZero); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(max_normal, func(max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_max_normal, + func(neg_max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + + EXPECT_FP_EQ_WITH_EXCEPTION(zero, + func(in.min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_zero, + func(in.neg_min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + + if (ForceRoundingMode r(RoundingMode::Downward); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(max_normal, func(max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_FP_EQ_WITH_EXCEPTION(-inf, func(neg_max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + + EXPECT_FP_EQ_WITH_EXCEPTION(zero, + func(in.min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_min_denormal, + func(in.neg_min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + + if (ForceRoundingMode r(RoundingMode::Upward); r.success) { + EXPECT_FP_EQ_WITH_EXCEPTION(inf, func(max_normal, neg_max_normal), + FE_OVERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_max_normal, + func(neg_max_normal, max_normal), + FE_OVERFLOW | FE_INEXACT); + + EXPECT_FP_EQ_WITH_EXCEPTION(min_denormal, + func(in.min_denormal, in.neg_min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + EXPECT_FP_EQ_WITH_EXCEPTION(neg_zero, + func(in.neg_min_denormal, in.min_denormal), + FE_UNDERFLOW | FE_INEXACT); + EXPECT_MATH_ERRNO(ERANGE); + } + } + + void test_inexact_results(SubFunc func) { + func(InType(1.0), min_denormal); + EXPECT_FP_EXCEPTION(FE_INEXACT); + } +}; + +#define LIST_SUB_TESTS(OutType, InType, func) \ + using LlvmLibcSubTest = SubTest<OutType, InType>; \ + TEST_F(LlvmLibcSubTest, SpecialNumbers) { test_special_numbers(&func); } \ + TEST_F(LlvmLibcSubTest, InvalidOperations) { \ + test_invalid_operations(&func); \ + } \ + TEST_F(LlvmLibcSubTest, RangeErrors) { test_range_errors(&func); } \ + TEST_F(LlvmLibcSubTest, InexactResults) { test_inexact_results(&func); } + +#endif // LLVM_LIBC_TEST_SRC_MATH_SMOKE_SUBTEST_H diff --git a/libc/test/src/math/smoke/f16addf_test.cpp b/libc/test/src/math/smoke/f16addf_test.cpp new file mode 100644 index 0000000..1e8b432 --- /dev/null +++ b/libc/test/src/math/smoke/f16addf_test.cpp @@ -0,0 +1,13 @@ +//===-- Unittests for f16addf ---------------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "AddTest.h" + +#include "src/math/f16addf.h" + +LIST_ADD_TESTS(float16, float, LIBC_NAMESPACE::f16addf) diff --git a/libc/test/src/math/smoke/f16subf_test.cpp b/libc/test/src/math/smoke/f16subf_test.cpp new file mode 100644 index 0000000..68ad948 --- /dev/null +++ b/libc/test/src/math/smoke/f16subf_test.cpp @@ -0,0 +1,13 @@ +//===-- Unittests for f16subf ---------------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "SubTest.h" + +#include "src/math/f16subf.h" + +LIST_SUB_TESTS(float16, float, LIBC_NAMESPACE::f16subf) diff --git a/libc/utils/MPFRWrapper/MPFRUtils.cpp b/libc/utils/MPFRWrapper/MPFRUtils.cpp index 379a631..f0a6538 100644 --- a/libc/utils/MPFRWrapper/MPFRUtils.cpp +++ b/libc/utils/MPFRWrapper/MPFRUtils.cpp @@ -185,6 +185,12 @@ public: return result; } + MPFRNumber add(const MPFRNumber &b) const { + MPFRNumber result(*this); + mpfr_add(result.value, value, b.value, mpfr_rounding); + return result; + } + MPFRNumber asin() const { MPFRNumber result(*this); mpfr_asin(result.value, value, mpfr_rounding); @@ -463,6 +469,12 @@ public: return result; } + MPFRNumber sub(const MPFRNumber &b) const { + MPFRNumber result(*this); + mpfr_sub(result.value, value, b.value, mpfr_rounding); + return result; + } + MPFRNumber tan() const { MPFRNumber result(*this); mpfr_tan(result.value, value, mpfr_rounding); @@ -487,6 +499,12 @@ public: return result; } + MPFRNumber fmul(const MPFRNumber &b) { + MPFRNumber result(*this); + mpfr_mul(result.value, value, b.value, mpfr_rounding); + return result; + } + cpp::string str() const { // 200 bytes should be more than sufficient to hold a 100-digit number // plus additional bytes for the decimal point, '-' sign etc. @@ -728,6 +746,8 @@ binary_operation_one_output(Operation op, InputType x, InputType y, MPFRNumber inputX(x, precision, rounding); MPFRNumber inputY(y, precision, rounding); switch (op) { + case Operation::Add: + return inputX.add(inputY); case Operation::Atan2: return inputX.atan2(inputY); case Operation::Div: @@ -738,6 +758,10 @@ binary_operation_one_output(Operation op, InputType x, InputType y, return inputX.hypot(inputY); case Operation::Pow: return inputX.pow(inputY); + case Operation::Sub: + return inputX.sub(inputY); + case Operation::Fmul: + return inputX.fmul(inputY); default: __builtin_unreachable(); } @@ -951,6 +975,9 @@ template void explain_binary_operation_one_output_error(Operation, const BinaryInput<long double> &, long double, double, RoundingMode); + +template void explain_binary_operation_one_output_error( + Operation, const BinaryInput<double> &, float, double, RoundingMode); #ifdef LIBC_TYPES_HAS_FLOAT16 template void explain_binary_operation_one_output_error( Operation, const BinaryInput<float16> &, float16, double, RoundingMode); @@ -1126,6 +1153,10 @@ template bool compare_binary_operation_one_output(Operation, template bool compare_binary_operation_one_output(Operation, const BinaryInput<long double> &, long double, double, RoundingMode); + +template bool compare_binary_operation_one_output(Operation, + const BinaryInput<double> &, + float, double, RoundingMode); #ifdef LIBC_TYPES_HAS_FLOAT16 template bool compare_binary_operation_one_output(Operation, const BinaryInput<float16> &, diff --git a/libc/utils/MPFRWrapper/MPFRUtils.h b/libc/utils/MPFRWrapper/MPFRUtils.h index 11e323b..213dc7a 100644 --- a/libc/utils/MPFRWrapper/MPFRUtils.h +++ b/libc/utils/MPFRWrapper/MPFRUtils.h @@ -71,11 +71,14 @@ enum class Operation : int { // input and produce a single floating point number of the same type as // output. BeginBinaryOperationsSingleOutput, + Add, Atan2, Div, Fmod, Hypot, Pow, + Sub, + Fmul, EndBinaryOperationsSingleOutput, // Operations which take two floating point numbers of the same type as @@ -237,7 +240,8 @@ public: bool is_silent() const override { return silent; } private: - template <typename T, typename U> bool match(T in, U out) { + template <typename InType, typename OutType> + bool match(InType in, OutType out) { return compare_unary_operation_single_output(op, in, out, ulp_tolerance, rounding); } @@ -259,13 +263,14 @@ private: rounding); } - template <typename T, typename U> - bool match(const TernaryInput<T> &in, U out) { + template <typename InType, typename OutType> + bool match(const TernaryInput<InType> &in, OutType out) { return compare_ternary_operation_one_output(op, in, out, ulp_tolerance, rounding); } - template <typename T, typename U> void explain_error(T in, U out) { + template <typename InType, typename OutType> + void explain_error(InType in, OutType out) { explain_unary_operation_single_output_error(op, in, out, ulp_tolerance, rounding); } @@ -287,8 +292,8 @@ private: rounding); } - template <typename T, typename U> - void explain_error(const TernaryInput<T> &in, U out) { + template <typename InType, typename OutType> + void explain_error(const TernaryInput<InType> &in, OutType out) { explain_ternary_operation_one_output_error(op, in, out, ulp_tolerance, rounding); } @@ -304,7 +309,9 @@ constexpr bool is_valid_operation() { (op == Operation::Sqrt && cpp::is_floating_point_v<InputType> && cpp::is_floating_point_v<OutputType> && sizeof(OutputType) <= sizeof(InputType)) || - (op == Operation::Div && internal::IsBinaryInput<InputType>::VALUE && + (Operation::BeginBinaryOperationsSingleOutput < op && + op < Operation::EndBinaryOperationsSingleOutput && + internal::IsBinaryInput<InputType>::VALUE && cpp::is_floating_point_v< typename internal::MakeScalarInput<InputType>::type> && cpp::is_floating_point_v<OutputType>) || diff --git a/libcxx/test/libcxx/utilities/utility/pairs/pairs.pair/pair.incomplete.compile.pass.cpp b/libcxx/test/libcxx/utilities/utility/pairs/pairs.pair/pair.incomplete.compile.pass.cpp new file mode 100644 index 0000000..16ee000 --- /dev/null +++ b/libcxx/test/libcxx/utilities/utility/pairs/pairs.pair/pair.incomplete.compile.pass.cpp @@ -0,0 +1,22 @@ +//===----------------------------------------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +// Check that instantiating pair doesn't look up type traits "too early", before +// the contained types have been completed. +// +// This is a regression test, to prevent a reoccurrance of the issue introduced +// in 5e1de27f680591a870d78e9952b23f76aed7f456. + +#include <utility> +#include <vector> + +struct Test { + std::vector<std::pair<int, Test> > v; +}; + +std::pair<int, Test> p; diff --git a/lldb/source/Commands/CMakeLists.txt b/lldb/source/Commands/CMakeLists.txt index 6a36c53..7639722 100644 --- a/lldb/source/Commands/CMakeLists.txt +++ b/lldb/source/Commands/CMakeLists.txt @@ -26,7 +26,7 @@ add_lldb_library(lldbCommands NO_PLUGIN_DEPENDENCIES CommandObjectQuit.cpp CommandObjectRegexCommand.cpp CommandObjectRegister.cpp - CommandObjectScript.cpp + CommandObjectScripting.cpp CommandObjectSession.cpp CommandObjectSettings.cpp CommandObjectSource.cpp diff --git a/lldb/source/Commands/CommandObjectScript.cpp b/lldb/source/Commands/CommandObjectScript.cpp deleted file mode 100644 index 25f25b8..0000000 --- a/lldb/source/Commands/CommandObjectScript.cpp +++ /dev/null @@ -1,113 +0,0 @@ -//===-- CommandObjectScript.cpp -------------------------------------------===// -// -// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. -// See https://llvm.org/LICENSE.txt for license information. -// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception -// -//===----------------------------------------------------------------------===// - -#include "CommandObjectScript.h" -#include "lldb/Core/Debugger.h" -#include "lldb/DataFormatters/DataVisualization.h" -#include "lldb/Host/Config.h" -#include "lldb/Host/OptionParser.h" -#include "lldb/Interpreter/CommandInterpreter.h" -#include "lldb/Interpreter/CommandOptionArgumentTable.h" -#include "lldb/Interpreter/CommandReturnObject.h" -#include "lldb/Interpreter/OptionArgParser.h" -#include "lldb/Interpreter/ScriptInterpreter.h" -#include "lldb/Utility/Args.h" - -using namespace lldb; -using namespace lldb_private; - -#define LLDB_OPTIONS_script -#include "CommandOptions.inc" - -Status CommandObjectScript::CommandOptions::SetOptionValue( - uint32_t option_idx, llvm::StringRef option_arg, - ExecutionContext *execution_context) { - Status error; - const int short_option = m_getopt_table[option_idx].val; - - switch (short_option) { - case 'l': - language = (lldb::ScriptLanguage)OptionArgParser::ToOptionEnum( - option_arg, GetDefinitions()[option_idx].enum_values, - eScriptLanguageNone, error); - if (!error.Success()) - error.SetErrorStringWithFormat("unrecognized value for language '%s'", - option_arg.str().c_str()); - break; - default: - llvm_unreachable("Unimplemented option"); - } - - return error; -} - -void CommandObjectScript::CommandOptions::OptionParsingStarting( - ExecutionContext *execution_context) { - language = lldb::eScriptLanguageNone; -} - -llvm::ArrayRef<OptionDefinition> -CommandObjectScript::CommandOptions::GetDefinitions() { - return llvm::ArrayRef(g_script_options); -} - -CommandObjectScript::CommandObjectScript(CommandInterpreter &interpreter) - : CommandObjectRaw( - interpreter, "script", - "Invoke the script interpreter with provided code and display any " - "results. Start the interactive interpreter if no code is supplied.", - "script [--language <scripting-language> --] [<script-code>]") {} - -CommandObjectScript::~CommandObjectScript() = default; - -void CommandObjectScript::DoExecute(llvm::StringRef command, - CommandReturnObject &result) { - // Try parsing the language option but when the command contains a raw part - // separated by the -- delimiter. - OptionsWithRaw raw_args(command); - if (raw_args.HasArgs()) { - if (!ParseOptions(raw_args.GetArgs(), result)) - return; - command = raw_args.GetRawPart(); - } - - lldb::ScriptLanguage language = - (m_options.language == lldb::eScriptLanguageNone) - ? m_interpreter.GetDebugger().GetScriptLanguage() - : m_options.language; - - if (language == lldb::eScriptLanguageNone) { - result.AppendError( - "the script-lang setting is set to none - scripting not available"); - return; - } - - ScriptInterpreter *script_interpreter = - GetDebugger().GetScriptInterpreter(true, language); - - if (script_interpreter == nullptr) { - result.AppendError("no script interpreter"); - return; - } - - // Script might change Python code we use for formatting. Make sure we keep - // up to date with it. - DataVisualization::ForceUpdate(); - - if (command.empty()) { - script_interpreter->ExecuteInterpreterLoop(); - result.SetStatus(eReturnStatusSuccessFinishNoResult); - return; - } - - // We can do better when reporting the status of one-liner script execution. - if (script_interpreter->ExecuteOneLine(command, &result)) - result.SetStatus(eReturnStatusSuccessFinishNoResult); - else - result.SetStatus(eReturnStatusFailed); -} diff --git a/lldb/source/Commands/CommandObjectScript.h b/lldb/source/Commands/CommandObjectScript.h deleted file mode 100644 index 3a8c4a8..0000000 --- a/lldb/source/Commands/CommandObjectScript.h +++ /dev/null @@ -1,42 +0,0 @@ -//===-- CommandObjectScript.h -----------------------------------*- C++ -*-===// -// -// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. -// See https://llvm.org/LICENSE.txt for license information. -// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception -// -//===----------------------------------------------------------------------===// - -#ifndef LLDB_SOURCE_INTERPRETER_COMMANDOBJECTSCRIPT_H -#define LLDB_SOURCE_INTERPRETER_COMMANDOBJECTSCRIPT_H - -#include "lldb/Interpreter/CommandObject.h" - -namespace lldb_private { - -class CommandObjectScript : public CommandObjectRaw { -public: - CommandObjectScript(CommandInterpreter &interpreter); - ~CommandObjectScript() override; - Options *GetOptions() override { return &m_options; } - - class CommandOptions : public Options { - public: - CommandOptions() = default; - ~CommandOptions() override = default; - Status SetOptionValue(uint32_t option_idx, llvm::StringRef option_arg, - ExecutionContext *execution_context) override; - void OptionParsingStarting(ExecutionContext *execution_context) override; - llvm::ArrayRef<OptionDefinition> GetDefinitions() override; - lldb::ScriptLanguage language = lldb::eScriptLanguageNone; - }; - -protected: - void DoExecute(llvm::StringRef command, CommandReturnObject &result) override; - -private: - CommandOptions m_options; -}; - -} // namespace lldb_private - -#endif // LLDB_SOURCE_INTERPRETER_COMMANDOBJECTSCRIPT_H diff --git a/lldb/source/Commands/CommandObjectScripting.cpp b/lldb/source/Commands/CommandObjectScripting.cpp new file mode 100644 index 0000000..fee0565 --- /dev/null +++ b/lldb/source/Commands/CommandObjectScripting.cpp @@ -0,0 +1,144 @@ +//===-- CommandObjectScripting.cpp ----------------------------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "CommandObjectScripting.h" +#include "lldb/Core/Debugger.h" +#include "lldb/DataFormatters/DataVisualization.h" +#include "lldb/Host/Config.h" +#include "lldb/Host/OptionParser.h" +#include "lldb/Interpreter/CommandInterpreter.h" +#include "lldb/Interpreter/CommandOptionArgumentTable.h" +#include "lldb/Interpreter/CommandReturnObject.h" +#include "lldb/Interpreter/OptionArgParser.h" +#include "lldb/Interpreter/ScriptInterpreter.h" +#include "lldb/Utility/Args.h" + +using namespace lldb; +using namespace lldb_private; + +#define LLDB_OPTIONS_scripting_run +#include "CommandOptions.inc" + +class CommandObjectScriptingRun : public CommandObjectRaw { +public: + CommandObjectScriptingRun(CommandInterpreter &interpreter) + : CommandObjectRaw( + interpreter, "scripting run", + "Invoke the script interpreter with provided code and display any " + "results. Start the interactive interpreter if no code is " + "supplied.", + "scripting run [--language <scripting-language> --] " + "[<script-code>]") {} + + ~CommandObjectScriptingRun() override = default; + + Options *GetOptions() override { return &m_options; } + + class CommandOptions : public Options { + public: + CommandOptions() = default; + ~CommandOptions() override = default; + Status SetOptionValue(uint32_t option_idx, llvm::StringRef option_arg, + ExecutionContext *execution_context) override { + Status error; + const int short_option = m_getopt_table[option_idx].val; + + switch (short_option) { + case 'l': + language = (lldb::ScriptLanguage)OptionArgParser::ToOptionEnum( + option_arg, GetDefinitions()[option_idx].enum_values, + eScriptLanguageNone, error); + if (!error.Success()) + error.SetErrorStringWithFormat("unrecognized value for language '%s'", + option_arg.str().c_str()); + break; + default: + llvm_unreachable("Unimplemented option"); + } + + return error; + } + + void OptionParsingStarting(ExecutionContext *execution_context) override { + language = lldb::eScriptLanguageNone; + } + + llvm::ArrayRef<OptionDefinition> GetDefinitions() override { + return llvm::ArrayRef(g_scripting_run_options); + } + + lldb::ScriptLanguage language = lldb::eScriptLanguageNone; + }; + +protected: + void DoExecute(llvm::StringRef command, + CommandReturnObject &result) override { + // Try parsing the language option but when the command contains a raw part + // separated by the -- delimiter. + OptionsWithRaw raw_args(command); + if (raw_args.HasArgs()) { + if (!ParseOptions(raw_args.GetArgs(), result)) + return; + command = raw_args.GetRawPart(); + } + + lldb::ScriptLanguage language = + (m_options.language == lldb::eScriptLanguageNone) + ? m_interpreter.GetDebugger().GetScriptLanguage() + : m_options.language; + + if (language == lldb::eScriptLanguageNone) { + result.AppendError( + "the script-lang setting is set to none - scripting not available"); + return; + } + + ScriptInterpreter *script_interpreter = + GetDebugger().GetScriptInterpreter(true, language); + + if (script_interpreter == nullptr) { + result.AppendError("no script interpreter"); + return; + } + + // Script might change Python code we use for formatting. Make sure we keep + // up to date with it. + DataVisualization::ForceUpdate(); + + if (command.empty()) { + script_interpreter->ExecuteInterpreterLoop(); + result.SetStatus(eReturnStatusSuccessFinishNoResult); + return; + } + + // We can do better when reporting the status of one-liner script execution. + if (script_interpreter->ExecuteOneLine(command, &result)) + result.SetStatus(eReturnStatusSuccessFinishNoResult); + else + result.SetStatus(eReturnStatusFailed); + } + +private: + CommandOptions m_options; +}; + +#pragma mark CommandObjectMultiwordScripting + +// CommandObjectMultiwordScripting + +CommandObjectMultiwordScripting::CommandObjectMultiwordScripting( + CommandInterpreter &interpreter) + : CommandObjectMultiword( + interpreter, "scripting", + "Commands for operating on the scripting functionnalities.", + "scripting <subcommand> [<subcommand-options>]") { + LoadSubCommand("run", + CommandObjectSP(new CommandObjectScriptingRun(interpreter))); +} + +CommandObjectMultiwordScripting::~CommandObjectMultiwordScripting() = default; diff --git a/lldb/source/Commands/CommandObjectScripting.h b/lldb/source/Commands/CommandObjectScripting.h new file mode 100644 index 0000000..0ccd579 --- /dev/null +++ b/lldb/source/Commands/CommandObjectScripting.h @@ -0,0 +1,25 @@ +//===-- CommandObjectScripting.h --------------------------------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLDB_SOURCE_INTERPRETER_COMMANDOBJECTSCRIPTING_H +#define LLDB_SOURCE_INTERPRETER_COMMANDOBJECTSCRIPTING_H + +#include "lldb/Interpreter/CommandObjectMultiword.h" + +namespace lldb_private { + +class CommandObjectMultiwordScripting : public CommandObjectMultiword { +public: + CommandObjectMultiwordScripting(CommandInterpreter &interpreter); + + ~CommandObjectMultiwordScripting() override; +}; + +} // namespace lldb_private + +#endif // LLDB_SOURCE_INTERPRETER_COMMANDOBJECTSCRIPTING_H diff --git a/lldb/source/Commands/Options.td b/lldb/source/Commands/Options.td index fa8af7c..24e97f3 100644 --- a/lldb/source/Commands/Options.td +++ b/lldb/source/Commands/Options.td @@ -835,7 +835,7 @@ let Command = "container add" in { Desc<"Overwrite an existing command at this node.">; } -let Command = "script" in { +let Command = "scripting run" in { def script_language : Option<"language", "l">, EnumArg<"ScriptLang">, Desc<"Specify the scripting " " language. If none is specific the default scripting language is used.">; diff --git a/lldb/source/Interpreter/CommandInterpreter.cpp b/lldb/source/Interpreter/CommandInterpreter.cpp index 40c915b..fc07168 100644 --- a/lldb/source/Interpreter/CommandInterpreter.cpp +++ b/lldb/source/Interpreter/CommandInterpreter.cpp @@ -33,7 +33,7 @@ #include "Commands/CommandObjectQuit.h" #include "Commands/CommandObjectRegexCommand.h" #include "Commands/CommandObjectRegister.h" -#include "Commands/CommandObjectScript.h" +#include "Commands/CommandObjectScripting.h" #include "Commands/CommandObjectSession.h" #include "Commands/CommandObjectSettings.h" #include "Commands/CommandObjectSource.h" @@ -518,6 +518,15 @@ void CommandInterpreter::Initialize() { AddAlias("re", cmd_obj_sp); } + cmd_obj_sp = GetCommandSPExact("scripting run"); + if (cmd_obj_sp) { + AddAlias("sc", cmd_obj_sp); + AddAlias("scr", cmd_obj_sp); + AddAlias("scri", cmd_obj_sp); + AddAlias("scrip", cmd_obj_sp); + AddAlias("script", cmd_obj_sp); + } + cmd_obj_sp = GetCommandSPExact("session history"); if (cmd_obj_sp) { AddAlias("history", cmd_obj_sp); @@ -569,7 +578,7 @@ void CommandInterpreter::LoadCommandDictionary() { REGISTER_COMMAND_OBJECT("process", CommandObjectMultiwordProcess); REGISTER_COMMAND_OBJECT("quit", CommandObjectQuit); REGISTER_COMMAND_OBJECT("register", CommandObjectRegister); - REGISTER_COMMAND_OBJECT("script", CommandObjectScript); + REGISTER_COMMAND_OBJECT("scripting", CommandObjectMultiwordScripting); REGISTER_COMMAND_OBJECT("settings", CommandObjectMultiwordSettings); REGISTER_COMMAND_OBJECT("session", CommandObjectSession); REGISTER_COMMAND_OBJECT("source", CommandObjectMultiwordSource); diff --git a/lldb/source/Plugins/ObjectFile/Mach-O/ObjectFileMachO.cpp b/lldb/source/Plugins/ObjectFile/Mach-O/ObjectFileMachO.cpp index 2979bf69..0dcb1be 100644 --- a/lldb/source/Plugins/ObjectFile/Mach-O/ObjectFileMachO.cpp +++ b/lldb/source/Plugins/ObjectFile/Mach-O/ObjectFileMachO.cpp @@ -6159,10 +6159,6 @@ Section *ObjectFileMachO::GetMachHeaderSection() { bool ObjectFileMachO::SectionIsLoadable(const Section *section) { if (!section) return false; - const bool is_dsym = (m_header.filetype == MH_DSYM); - if (section->GetFileSize() == 0 && !is_dsym && - section->GetName() != GetSegmentNameDATA()) - return false; if (section->IsThreadSpecific()) return false; if (GetModule().get() != section->GetModule().get()) @@ -6202,6 +6198,7 @@ lldb::addr_t ObjectFileMachO::CalculateSectionLoadAddressForMemoryImage( bool ObjectFileMachO::SetLoadAddress(Target &target, lldb::addr_t value, bool value_is_offset) { + Log *log(GetLog(LLDBLog::DynamicLoader)); ModuleSP module_sp = GetModule(); if (!module_sp) return false; @@ -6217,17 +6214,33 @@ bool ObjectFileMachO::SetLoadAddress(Target &target, lldb::addr_t value, // malformed. const bool warn_multiple = true; + if (log) { + StreamString logmsg; + logmsg << "ObjectFileMachO::SetLoadAddress "; + if (GetFileSpec()) + logmsg << "path='" << GetFileSpec().GetPath() << "' "; + if (GetUUID()) { + logmsg << "uuid=" << GetUUID().GetAsString(); + } + LLDB_LOGF(log, "%s", logmsg.GetData()); + } if (value_is_offset) { // "value" is an offset to apply to each top level segment for (size_t sect_idx = 0; sect_idx < num_sections; ++sect_idx) { // Iterate through the object file sections to find all of the // sections that size on disk (to avoid __PAGEZERO) and load them SectionSP section_sp(section_list->GetSectionAtIndex(sect_idx)); - if (SectionIsLoadable(section_sp.get())) + if (SectionIsLoadable(section_sp.get())) { + LLDB_LOGF(log, + "ObjectFileMachO::SetLoadAddress segment '%s' load addr is " + "0x%" PRIx64, + section_sp->GetName().AsCString(), + section_sp->GetFileAddress() + value); if (target.GetSectionLoadList().SetSectionLoadAddress( section_sp, section_sp->GetFileAddress() + value, warn_multiple)) ++num_loaded_sections; + } } } else { // "value" is the new base address of the mach_header, adjust each @@ -6242,6 +6255,10 @@ bool ObjectFileMachO::SetLoadAddress(Target &target, lldb::addr_t value, CalculateSectionLoadAddressForMemoryImage( value, mach_header_section, section_sp.get()); if (section_load_addr != LLDB_INVALID_ADDRESS) { + LLDB_LOGF(log, + "ObjectFileMachO::SetLoadAddress segment '%s' load addr is " + "0x%" PRIx64, + section_sp->GetName().AsCString(), section_load_addr); if (target.GetSectionLoadList().SetSectionLoadAddress( section_sp, section_load_addr, warn_multiple)) ++num_loaded_sections; diff --git a/lldb/source/Utility/Listener.cpp b/lldb/source/Utility/Listener.cpp index 6a74c53..5aacb41 100644 --- a/lldb/source/Utility/Listener.cpp +++ b/lldb/source/Utility/Listener.cpp @@ -30,7 +30,7 @@ Listener::Listener(const char *name) Listener::~Listener() { Log *log = GetLog(LLDBLog::Object); - Clear(); + // Don't call Clear() from here as that can cause races. See #96750. LLDB_LOGF(log, "%p Listener::%s('%s')", static_cast<void *>(this), __FUNCTION__, m_name.c_str()); diff --git a/lldb/test/API/commands/settings/use_source_cache/TestUseSourceCache.py b/lldb/test/API/commands/settings/use_source_cache/TestUseSourceCache.py index c54345a..4215990 100644 --- a/lldb/test/API/commands/settings/use_source_cache/TestUseSourceCache.py +++ b/lldb/test/API/commands/settings/use_source_cache/TestUseSourceCache.py @@ -18,7 +18,7 @@ class SettingsUseSourceCacheTestCase(TestBase): self.set_use_source_cache_and_test(False) @skipIf(hostoslist=no_match(["windows"])) - @skipIf(oslist=["windows"], archs=["aarch64"]) # Fails on windows 11 + @skipIf(oslist=["windows"]) # Fails on windows 11 def test_set_use_source_cache_true(self): """Test that after 'set use-source-cache false', files are locked.""" self.set_use_source_cache_and_test(True) diff --git a/lldb/test/API/functionalities/abbreviation/TestAbbreviations.py b/lldb/test/API/functionalities/abbreviation/TestAbbreviations.py index 10431e4..02ee581 100644 --- a/lldb/test/API/functionalities/abbreviation/TestAbbreviations.py +++ b/lldb/test/API/functionalities/abbreviation/TestAbbreviations.py @@ -80,7 +80,13 @@ class AbbreviationsTestCase(TestBase): # Check a command that wants the raw input. command_interpreter.ResolveCommand(r"""sc print("\n\n\tHello!\n")""", result) self.assertTrue(result.Succeeded()) - self.assertEqual(r"""script print("\n\n\tHello!\n")""", result.GetOutput()) + self.assertEqual( + r"""scripting run print("\n\n\tHello!\n")""", result.GetOutput() + ) + + command_interpreter.ResolveCommand("script 1+1", result) + self.assertTrue(result.Succeeded()) + self.assertEqual("scripting run 1+1", result.GetOutput()) # Prompt changing stuff should be tested, but this doesn't seem like the # right test to do it in. It has nothing to do with aliases or abbreviations. diff --git a/lldb/test/Shell/SymbolFile/DWARF/x86/dwp-foreign-type-units.cpp b/lldb/test/Shell/SymbolFile/DWARF/x86/dwp-foreign-type-units.cpp index 4df1b33..ef15d41 100644 --- a/lldb/test/Shell/SymbolFile/DWARF/x86/dwp-foreign-type-units.cpp +++ b/lldb/test/Shell/SymbolFile/DWARF/x86/dwp-foreign-type-units.cpp @@ -1,4 +1,6 @@ // REQUIRES: lld +// Is flaky on Windows. +// UNSUPPORTED: system-windows // This test will make a type that will be compiled differently into two // different .dwo files in a type unit with the same type hash, but with diff --git a/llvm/CMakeLists.txt b/llvm/CMakeLists.txt index 91a2b61..1261896 100644 --- a/llvm/CMakeLists.txt +++ b/llvm/CMakeLists.txt @@ -1135,25 +1135,6 @@ configure_file( ${LLVM_MAIN_INCLUDE_DIR}/llvm/Config/abi-breaking.h.cmake ${LLVM_INCLUDE_DIR}/llvm/Config/abi-breaking.h) -# Add target for generating source rpm package. -set(LLVM_SRPM_USER_BINARY_SPECFILE ${CMAKE_CURRENT_SOURCE_DIR}/llvm.spec.in - CACHE FILEPATH ".spec file to use for srpm generation") -set(LLVM_SRPM_BINARY_SPECFILE ${CMAKE_CURRENT_BINARY_DIR}/llvm.spec) -set(LLVM_SRPM_DIR "${CMAKE_CURRENT_BINARY_DIR}/srpm") - -get_source_info(${CMAKE_CURRENT_SOURCE_DIR} revision repository) -string(LENGTH "${revision}" revision_length) -set(LLVM_RPM_SPEC_REVISION "${revision}") - -configure_file( - ${LLVM_SRPM_USER_BINARY_SPECFILE} - ${LLVM_SRPM_BINARY_SPECFILE} @ONLY) - -add_custom_target(srpm - COMMAND cpack -G TGZ --config CPackSourceConfig.cmake -B ${LLVM_SRPM_DIR}/SOURCES - COMMAND rpmbuild -bs --define '_topdir ${LLVM_SRPM_DIR}' ${LLVM_SRPM_BINARY_SPECFILE}) -set_target_properties(srpm PROPERTIES FOLDER "LLVM/Misc") - if(APPLE AND DARWIN_LTO_LIBRARY) set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -Wl,-lto_library -Wl,${DARWIN_LTO_LIBRARY}") diff --git a/llvm/docs/CommandGuide/FileCheck.rst b/llvm/docs/CommandGuide/FileCheck.rst index 432dafa..af49304 100644 --- a/llvm/docs/CommandGuide/FileCheck.rst +++ b/llvm/docs/CommandGuide/FileCheck.rst @@ -731,7 +731,7 @@ The first check line matches a regex ``%[a-z]+`` and captures it into the string variable ``REGISTER``. The second line verifies that whatever is in ``REGISTER`` occurs later in the file after an "``andw``". :program:`FileCheck` string substitution blocks are always contained in ``[[ ]]`` pairs, and string -variable names can be formed with the regex ``[a-zA-Z_][a-zA-Z0-9_]*``. If a +variable names can be formed with the regex ``\$[a-zA-Z_][a-zA-Z0-9_]*``. If a colon follows the name, then it is a definition of the variable; otherwise, it is a substitution. diff --git a/llvm/docs/LangRef.rst b/llvm/docs/LangRef.rst index 211fee5..e2c4720 100644 --- a/llvm/docs/LangRef.rst +++ b/llvm/docs/LangRef.rst @@ -5381,6 +5381,7 @@ NVPTX: - ``c`` or ``h``: A 16-bit integer register. - ``r``: A 32-bit integer register. - ``l`` or ``N``: A 64-bit integer register. +- ``q``: A 128-bit integer register. - ``f``: A 32-bit float register. - ``d``: A 64-bit float register. diff --git a/llvm/include/llvm/IR/IRBuilder.h b/llvm/include/llvm/IR/IRBuilder.h index f81cfc0..31a1fef 100644 --- a/llvm/include/llvm/IR/IRBuilder.h +++ b/llvm/include/llvm/IR/IRBuilder.h @@ -1023,6 +1023,14 @@ public: Name); } + /// Create call to the ldexp intrinsic. + Value *CreateLdexp(Value *Src, Value *Exp, Instruction *FMFSource = nullptr, + const Twine &Name = "") { + assert(!IsFPConstrained && "TODO: Support strictfp"); + return CreateIntrinsic(Intrinsic::ldexp, {Src->getType(), Exp->getType()}, + {Src, Exp}, FMFSource, Name); + } + /// Create a call to the arithmetic_fence intrinsic. CallInst *CreateArithmeticFence(Value *Val, Type *DstType, const Twine &Name = "") { diff --git a/llvm/include/llvm/IR/PassManagerImpl.h b/llvm/include/llvm/IR/PassManagerImpl.h index 7887719..67e3fbe 100644 --- a/llvm/include/llvm/IR/PassManagerImpl.h +++ b/llvm/include/llvm/IR/PassManagerImpl.h @@ -30,20 +30,24 @@ PreservedAnalyses PassManager<IRUnitT, AnalysisManagerT, ExtraArgTs...>::run( IRUnitT &IR, AnalysisManagerT &AM, ExtraArgTs... ExtraArgs) { class StackTraceEntry : public PrettyStackTraceEntry { const PassInstrumentation &PI; - PassConceptT &Pass; IRUnitT &IR; + PassConceptT *Pass = nullptr; public: - explicit StackTraceEntry(const PassInstrumentation &PI, PassConceptT &Pass, - IRUnitT &IR) - : PI(PI), Pass(Pass), IR(IR) {} + explicit StackTraceEntry(const PassInstrumentation &PI, IRUnitT &IR) + : PI(PI), IR(IR) {} + + void setPass(PassConceptT *P) { Pass = P; } void print(raw_ostream &OS) const override { OS << "Running pass \""; - Pass.printPipeline(OS, [this](StringRef ClassName) { - auto PassName = PI.getPassNameForClassName(ClassName); - return PassName.empty() ? ClassName : PassName; - }); + if (Pass) + Pass->printPipeline(OS, [this](StringRef ClassName) { + auto PassName = PI.getPassNameForClassName(ClassName); + return PassName.empty() ? ClassName : PassName; + }); + else + OS << "unknown"; OS << "\" on "; printIRUnitNameForStackTrace(OS, IR); OS << "\n"; @@ -64,14 +68,16 @@ PreservedAnalyses PassManager<IRUnitT, AnalysisManagerT, ExtraArgTs...>::run( // for duration of these passes. ScopedDbgInfoFormatSetter FormatSetter(IR, UseNewDbgInfoFormat); + StackTraceEntry Entry(PI, IR); for (auto &Pass : Passes) { + Entry.setPass(&*Pass); + // Check the PassInstrumentation's BeforePass callbacks before running the // pass, skip its execution completely if asked to (callback returns // false). if (!PI.runBeforePass<IRUnitT>(*Pass, IR)) continue; - StackTraceEntry Entry(PI, *Pass, IR); PreservedAnalyses PassPA = Pass->run(IR, AM, ExtraArgs...); // Update the analysis manager as each pass runs and potentially diff --git a/llvm/include/llvm/MC/MCAsmBackend.h b/llvm/include/llvm/MC/MCAsmBackend.h index 7841e8a..f91c8e1 100644 --- a/llvm/include/llvm/MC/MCAsmBackend.h +++ b/llvm/include/llvm/MC/MCAsmBackend.h @@ -24,7 +24,6 @@ class MCFragment; class MCLEBFragment; class MCRelaxableFragment; class MCSymbol; -class MCAsmLayout; class MCAssembler; class MCContext; struct MCDwarfFrameInfo; @@ -219,8 +218,7 @@ public: const MCSubtargetInfo *STI) const = 0; /// Give backend an opportunity to finish layout after relaxation - virtual void finishLayout(MCAssembler const &Asm, - MCAsmLayout &Layout) const {} + virtual void finishLayout(MCAssembler const &Asm) const {} /// Handle any target-specific assembler flags. By default, do nothing. virtual void handleAssemblerFlag(MCAssemblerFlag Flag) {} diff --git a/llvm/include/llvm/MC/MCAsmLayout.h b/llvm/include/llvm/MC/MCAsmLayout.h deleted file mode 100644 index 765cc1e..0000000 --- a/llvm/include/llvm/MC/MCAsmLayout.h +++ /dev/null @@ -1,31 +0,0 @@ -//===- MCAsmLayout.h - Assembly Layout Object -------------------*- C++ -*-===// -// -// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. -// See https://llvm.org/LICENSE.txt for license information. -// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception -// -//===----------------------------------------------------------------------===// - -#ifndef LLVM_MC_MCASMLAYOUT_H -#define LLVM_MC_MCASMLAYOUT_H - -#include "llvm/ADT/DenseMap.h" -#include "llvm/ADT/SmallVector.h" - -namespace llvm { -class MCAssembler; -class MCSection; - -class MCAsmLayout { - MCAssembler &Assembler; - -public: - MCAsmLayout(MCAssembler &Assembler); - - /// Get the assembler object this is a layout for. - MCAssembler &getAssembler() const { return Assembler; } -}; - -} // end namespace llvm - -#endif diff --git a/llvm/include/llvm/MC/MCAssembler.h b/llvm/include/llvm/MC/MCAssembler.h index cc0267e..9cd65d3 100644 --- a/llvm/include/llvm/MC/MCAssembler.h +++ b/llvm/include/llvm/MC/MCAssembler.h @@ -46,7 +46,6 @@ class MCRelaxableFragment; class MCSymbolRefExpr; class raw_ostream; class MCAsmBackend; -class MCAsmLayout; class MCContext; class MCCodeEmitter; class MCFragment; @@ -71,8 +70,6 @@ struct DataRegionData { }; class MCAssembler { - friend class MCAsmLayout; - public: using SectionListType = std::vector<MCSection *>; using SymbolDataListType = std::vector<const MCSymbol *>; @@ -118,7 +115,7 @@ private: std::unique_ptr<MCCodeEmitter> Emitter; std::unique_ptr<MCObjectWriter> Writer; - MCAsmLayout *Layout = nullptr; + bool HasLayout = false; bool RelaxAll = false; bool SubsectionsViaSymbols = false; bool IncrementalLinkerCompatible = false; @@ -210,9 +207,6 @@ private: bool relaxCVDefRange(MCCVDefRangeFragment &DF); bool relaxPseudoProbeAddr(MCPseudoProbeAddrFragment &DF); - /// finishLayout - Finalize a layout, including fragment lowering. - void finishLayout(MCAsmLayout &Layout); - std::tuple<MCValue, uint64_t, bool> handleFixup(MCFragment &F, const MCFixup &Fixup, const MCSubtargetInfo *STI); @@ -346,7 +340,7 @@ public: void Finish(); // Layout all section and prepare them for emission. - void layout(MCAsmLayout &Layout); + void layout(); // FIXME: This does not belong here. bool getSubsectionsViaSymbols() const { return SubsectionsViaSymbols; } @@ -359,8 +353,7 @@ public: IncrementalLinkerCompatible = Value; } - MCAsmLayout *getLayout() const { return Layout; } - bool hasLayout() const { return Layout; } + bool hasLayout() const { return HasLayout; } bool getRelaxAll() const { return RelaxAll; } void setRelaxAll(bool Value) { RelaxAll = Value; } diff --git a/llvm/include/llvm/MC/MCCodeView.h b/llvm/include/llvm/MC/MCCodeView.h index b1d8fe3..3092327 100644 --- a/llvm/include/llvm/MC/MCCodeView.h +++ b/llvm/include/llvm/MC/MCCodeView.h @@ -22,7 +22,7 @@ #include <vector> namespace llvm { -class MCAsmLayout; +class MCAssembler; class MCCVDefRangeFragment; class MCCVInlineLineTableFragment; class MCDataFragment; @@ -199,7 +199,7 @@ public: const MCSymbol *FnEndSym); /// Encodes the binary annotations once we have a layout. - void encodeInlineLineTable(MCAsmLayout &Layout, + void encodeInlineLineTable(const MCAssembler &Asm, MCCVInlineLineTableFragment &F); MCFragment * @@ -207,7 +207,7 @@ public: ArrayRef<std::pair<const MCSymbol *, const MCSymbol *>> Ranges, StringRef FixedSizePortion); - void encodeDefRange(MCAsmLayout &Layout, MCCVDefRangeFragment &F); + void encodeDefRange(const MCAssembler &Asm, MCCVDefRangeFragment &F); /// Emits the string table substream. void emitStringTable(MCObjectStreamer &OS); diff --git a/llvm/include/llvm/MC/MCExpr.h b/llvm/include/llvm/MC/MCExpr.h index 40319dc..118b1dd 100644 --- a/llvm/include/llvm/MC/MCExpr.h +++ b/llvm/include/llvm/MC/MCExpr.h @@ -16,7 +16,6 @@ namespace llvm { class MCAsmInfo; -class MCAsmLayout; class MCAssembler; class MCContext; class MCFixup; @@ -54,7 +53,6 @@ private: SMLoc Loc; bool evaluateAsAbsolute(int64_t &Res, const MCAssembler *Asm, - const MCAsmLayout *Layout, const SectionAddrMap *Addrs, bool InSet) const; protected: @@ -101,18 +99,20 @@ public: bool evaluateAsAbsolute(int64_t &Res) const; bool evaluateAsAbsolute(int64_t &Res, const MCAssembler &Asm) const; bool evaluateAsAbsolute(int64_t &Res, const MCAssembler *Asm) const; - bool evaluateAsAbsolute(int64_t &Res, const MCAsmLayout &Layout) const; - bool evaluateKnownAbsolute(int64_t &Res, const MCAsmLayout &Layout) const; + /// Aggressive variant of evaluateAsRelocatable when relocations are + /// unavailable (e.g. .fill). Expects callers to handle errors when true is + /// returned. + bool evaluateKnownAbsolute(int64_t &Res, const MCAssembler &Asm) const; /// Try to evaluate the expression to a relocatable value, i.e. an /// expression of the fixed form (a - b + constant). /// /// \param Res - The relocatable value, if evaluation succeeds. - /// \param Layout - The assembler layout object to use for evaluating values. + /// \param Asm - The assembler object to use for evaluating values. /// \param Fixup - The Fixup object if available. /// \return - True on success. - bool evaluateAsRelocatable(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatable(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const; /// Try to evaluate the expression to the form (a - b + constant) where @@ -658,8 +658,7 @@ protected: public: virtual void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const = 0; - virtual bool evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + virtual bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const = 0; // allow Target Expressions to be checked for equality virtual bool isEqualTo(const MCExpr *x) const { return false; } diff --git a/llvm/include/llvm/MC/MCFragment.h b/llvm/include/llvm/MC/MCFragment.h index c8832e6..ee93cb9 100644 --- a/llvm/include/llvm/MC/MCFragment.h +++ b/llvm/include/llvm/MC/MCFragment.h @@ -30,7 +30,6 @@ class MCSubtargetInfo; class MCSymbol; class MCFragment { - friend class MCAsmLayout; friend class MCAssembler; friend class MCObjectStreamer; friend class MCSection; diff --git a/llvm/include/llvm/MC/MCLinkerOptimizationHint.h b/llvm/include/llvm/MC/MCLinkerOptimizationHint.h index b91fbc6..530f288 100644 --- a/llvm/include/llvm/MC/MCLinkerOptimizationHint.h +++ b/llvm/include/llvm/MC/MCLinkerOptimizationHint.h @@ -25,7 +25,7 @@ namespace llvm { class MachObjectWriter; -class MCAsmLayout; +class MCAssembler; class MCSymbol; class raw_ostream; @@ -108,8 +108,8 @@ class MCLOHDirective { /// Emit this directive in \p OutStream using the information available /// in the given \p ObjWriter and \p Layout to get the address of the /// arguments within the object file. - void emit_impl(raw_ostream &OutStream, const MachObjectWriter &ObjWriter, - const MCAsmLayout &Layout) const; + void emit_impl(const MCAssembler &Asm, raw_ostream &OutStream, + const MachObjectWriter &ObjWriter) const; public: using LOHArgs = SmallVectorImpl<MCSymbol *>; @@ -125,12 +125,12 @@ public: /// Emit this directive as: /// <kind, numArgs, addr1, ..., addrN> - void emit(MachObjectWriter &ObjWriter, const MCAsmLayout &Layout) const; + void emit(const MCAssembler &Asm, MachObjectWriter &ObjWriter) const; /// Get the size in bytes of this directive if emitted in \p ObjWriter with /// the given \p Layout. - uint64_t getEmitSize(const MachObjectWriter &ObjWriter, - const MCAsmLayout &Layout) const; + uint64_t getEmitSize(const MCAssembler &Asm, + const MachObjectWriter &ObjWriter) const; }; class MCLOHContainer { @@ -157,20 +157,20 @@ public: } /// Get the size of the directives if emitted. - uint64_t getEmitSize(const MachObjectWriter &ObjWriter, - const MCAsmLayout &Layout) const { + uint64_t getEmitSize(const MCAssembler &Asm, + const MachObjectWriter &ObjWriter) const { if (!EmitSize) { for (const MCLOHDirective &D : Directives) - EmitSize += D.getEmitSize(ObjWriter, Layout); + EmitSize += D.getEmitSize(Asm, ObjWriter); } return EmitSize; } /// Emit all Linker Optimization Hint in one big table. /// Each line of the table is emitted by LOHDirective::emit. - void emit(MachObjectWriter &ObjWriter, const MCAsmLayout &Layout) const { + void emit(const MCAssembler &Asm, MachObjectWriter &ObjWriter) const { for (const MCLOHDirective &D : Directives) - D.emit(ObjWriter, Layout); + D.emit(Asm, ObjWriter); } void reset() { diff --git a/llvm/include/llvm/MC/MCMachObjectWriter.h b/llvm/include/llvm/MC/MCMachObjectWriter.h index c685e27..172dadb 100644 --- a/llvm/include/llvm/MC/MCMachObjectWriter.h +++ b/llvm/include/llvm/MC/MCMachObjectWriter.h @@ -161,7 +161,7 @@ public: uint64_t getSectionAddress(const MCSection *Sec) const { return SectionAddress.lookup(Sec); } - uint64_t getSymbolAddress(const MCSymbol &S, const MCAsmLayout &Layout) const; + uint64_t getSymbolAddress(const MCSymbol &S, const MCAssembler &Asm) const; uint64_t getFragmentAddress(const MCAssembler &Asm, const MCFragment *Fragment) const; @@ -212,7 +212,7 @@ public: uint32_t FirstUndefinedSymbol, uint32_t NumUndefinedSymbols, uint32_t IndirectSymbolOffset, uint32_t NumIndirectSymbols); - void writeNlist(MachSymbolData &MSD, const MCAsmLayout &Layout); + void writeNlist(MachSymbolData &MSD, const MCAssembler &Asm); void writeLinkeditLoadCommand(uint32_t Type, uint32_t DataOffset, uint32_t DataSize); diff --git a/llvm/include/llvm/MC/MCObjectWriter.h b/llvm/include/llvm/MC/MCObjectWriter.h index 4184c4f..0111446 100644 --- a/llvm/include/llvm/MC/MCObjectWriter.h +++ b/llvm/include/llvm/MC/MCObjectWriter.h @@ -15,7 +15,6 @@ namespace llvm { -class MCAsmLayout; class MCAssembler; class MCFixup; class MCFragment; diff --git a/llvm/include/llvm/Support/LogicalResult.h b/llvm/include/llvm/Support/LogicalResult.h new file mode 100644 index 0000000..2cfd866b --- /dev/null +++ b/llvm/include/llvm/Support/LogicalResult.h @@ -0,0 +1,128 @@ +//===- LogicalResult.h - Utilities for handling success/failure -*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_SUPPORT_LOGICALRESULT_H +#define LLVM_SUPPORT_LOGICALRESULT_H + +#include <cassert> +#include <optional> + +namespace llvm { +/// This class represents an efficient way to signal success or failure. It +/// should be preferred over the use of `bool` when appropriate, as it avoids +/// all of the ambiguity that arises in interpreting a boolean result. This +/// class is marked as NODISCARD to ensure that the result is processed. Users +/// may explicitly discard a result by using `(void)`, e.g. +/// `(void)functionThatReturnsALogicalResult();`. Given the intended nature of +/// this class, it generally shouldn't be used as the result of functions that +/// very frequently have the result ignored. This class is intended to be used +/// in conjunction with the utility functions below. +struct [[nodiscard]] LogicalResult { +public: + /// If isSuccess is true a `success` result is generated, otherwise a + /// 'failure' result is generated. + static LogicalResult success(bool IsSuccess = true) { + return LogicalResult(IsSuccess); + } + + /// If isFailure is true a `failure` result is generated, otherwise a + /// 'success' result is generated. + static LogicalResult failure(bool IsFailure = true) { + return LogicalResult(!IsFailure); + } + + /// Returns true if the provided LogicalResult corresponds to a success value. + constexpr bool succeeded() const { return IsSuccess; } + + /// Returns true if the provided LogicalResult corresponds to a failure value. + constexpr bool failed() const { return !IsSuccess; } + +private: + LogicalResult(bool IsSuccess) : IsSuccess(IsSuccess) {} + + /// Boolean indicating if this is a success result, if false this is a + /// failure result. + bool IsSuccess; +}; + +/// Utility function to generate a LogicalResult. If isSuccess is true a +/// `success` result is generated, otherwise a 'failure' result is generated. +inline LogicalResult success(bool IsSuccess = true) { + return LogicalResult::success(IsSuccess); +} + +/// Utility function to generate a LogicalResult. If isFailure is true a +/// `failure` result is generated, otherwise a 'success' result is generated. +inline LogicalResult failure(bool IsFailure = true) { + return LogicalResult::failure(IsFailure); +} + +/// Utility function that returns true if the provided LogicalResult corresponds +/// to a success value. +inline bool succeeded(LogicalResult Result) { return Result.succeeded(); } + +/// Utility function that returns true if the provided LogicalResult corresponds +/// to a failure value. +inline bool failed(LogicalResult Result) { return Result.failed(); } + +/// This class provides support for representing a failure result, or a valid +/// value of type `T`. This allows for integrating with LogicalResult, while +/// also providing a value on the success path. +template <typename T> class [[nodiscard]] FailureOr : public std::optional<T> { +public: + /// Allow constructing from a LogicalResult. The result *must* be a failure. + /// Success results should use a proper instance of type `T`. + FailureOr(LogicalResult Result) { + assert(failed(Result) && + "success should be constructed with an instance of 'T'"); + } + FailureOr() : FailureOr(failure()) {} + FailureOr(T &&Y) : std::optional<T>(std::forward<T>(Y)) {} + FailureOr(const T &Y) : std::optional<T>(Y) {} + template <typename U, + std::enable_if_t<std::is_constructible<T, U>::value> * = nullptr> + FailureOr(const FailureOr<U> &Other) + : std::optional<T>(failed(Other) ? std::optional<T>() + : std::optional<T>(*Other)) {} + + operator LogicalResult() const { return success(has_value()); } + +private: + /// Hide the bool conversion as it easily creates confusion. + using std::optional<T>::operator bool; + using std::optional<T>::has_value; +}; + +/// Wrap a value on the success path in a FailureOr of the same value type. +template <typename T, + typename = std::enable_if_t<!std::is_convertible_v<T, bool>>> +inline auto success(T &&Y) { + return FailureOr<std::decay_t<T>>(std::forward<T>(Y)); +} + +/// This class represents success/failure for parsing-like operations that find +/// it important to chain together failable operations with `||`. This is an +/// extended version of `LogicalResult` that allows for explicit conversion to +/// bool. +/// +/// This class should not be used for general error handling cases - we prefer +/// to keep the logic explicit with the `succeeded`/`failed` predicates. +/// However, traditional monadic-style parsing logic can sometimes get +/// swallowed up in boilerplate without this, so we provide this for narrow +/// cases where it is important. +/// +class [[nodiscard]] ParseResult : public LogicalResult { +public: + ParseResult(LogicalResult Result = success()) : LogicalResult(Result) {} + + /// Failure is true in a boolean context. + constexpr explicit operator bool() const { return failed(); } +}; +} // namespace llvm + +#endif // LLVM_SUPPORT_LOGICALRESULT_H diff --git a/llvm/include/llvm/TargetParser/RISCVISAInfo.h b/llvm/include/llvm/TargetParser/RISCVISAInfo.h index 12f6b46..5e9cf67 100644 --- a/llvm/include/llvm/TargetParser/RISCVISAInfo.h +++ b/llvm/include/llvm/TargetParser/RISCVISAInfo.h @@ -87,8 +87,6 @@ private: RISCVISAUtils::OrderedExtensionMap Exts; - bool addExtension(StringRef ExtName, RISCVISAUtils::ExtensionVersion Version); - Error checkDependency(); void updateImplication(); diff --git a/llvm/lib/Analysis/InlineSizeEstimatorAnalysis.cpp b/llvm/lib/Analysis/InlineSizeEstimatorAnalysis.cpp index 37842b1..4074b67 100644 --- a/llvm/lib/Analysis/InlineSizeEstimatorAnalysis.cpp +++ b/llvm/lib/Analysis/InlineSizeEstimatorAnalysis.cpp @@ -30,7 +30,6 @@ AnalysisKey InlineSizeEstimatorAnalysis::Key; #include "llvm/IR/BasicBlock.h" #include "llvm/IR/Dominators.h" #include "llvm/IR/Instructions.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/Support/Casting.h" #include "llvm/Support/CommandLine.h" #include <algorithm> diff --git a/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp b/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp index 0033f0a..c52cbff 100644 --- a/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp +++ b/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp @@ -1128,7 +1128,7 @@ static void emitKill(const MachineInstr *MI, AsmPrinter &AP) { OS << ' ' << (Op.isDef() ? "def " : "killed ") << printReg(Op.getReg(), AP.MF->getSubtarget().getRegisterInfo()); } - AP.OutStreamer->AddComment(OS.str()); + AP.OutStreamer->AddComment(Str); AP.OutStreamer->addBlankLine(); } @@ -1235,7 +1235,7 @@ static bool emitDebugValueComment(const MachineInstr *MI, AsmPrinter &AP) { } // NOTE: Want this comment at start of line, don't emit with AddComment. - AP.OutStreamer->emitRawComment(OS.str()); + AP.OutStreamer->emitRawComment(Str); return true; } @@ -2274,7 +2274,7 @@ void AsmPrinter::emitRemarksSection(remarks::RemarkStreamer &RS) { OutContext.getObjectFileInfo()->getRemarksSection(); OutStreamer->switchSection(RemarksSection); - OutStreamer->emitBinaryData(OS.str()); + OutStreamer->emitBinaryData(Buf); } bool AsmPrinter::doFinalization(Module &M) { @@ -3282,7 +3282,7 @@ const MCExpr *AsmPrinter::lowerConstant(const Constant *CV) { OS << "Unsupported expression in static initializer: "; CE->printAsOperand(OS, /*PrintType=*/false, !MF ? nullptr : MF->getFunction().getParent()); - report_fatal_error(Twine(OS.str())); + report_fatal_error(Twine(S)); } static void emitGlobalConstantImpl(const DataLayout &DL, const Constant *C, diff --git a/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp b/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp index 08e3c20..5a7013c 100644 --- a/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp +++ b/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp @@ -311,7 +311,7 @@ static void EmitInlineAsmStr(const char *AsmStr, const MachineInstr *MI, std::string msg; raw_string_ostream Msg(msg); Msg << "invalid operand in inline asm: '" << AsmStr << "'"; - MMI->getModule()->getContext().emitError(LocCookie, Msg.str()); + MMI->getModule()->getContext().emitError(LocCookie, msg); } } break; @@ -411,7 +411,7 @@ void AsmPrinter::emitInlineAsm(const MachineInstr *MI) const { } } - emitInlineAsm(OS.str(), getSubtargetInfo(), TM.Options.MCOptions, LocMD, + emitInlineAsm(StringData, getSubtargetInfo(), TM.Options.MCOptions, LocMD, MI->getInlineAsmDialect()); // Emit the #NOAPP end marker. This has to happen even if verbose-asm isn't diff --git a/llvm/lib/CodeGen/GlobalISel/IRTranslator.cpp b/llvm/lib/CodeGen/GlobalISel/IRTranslator.cpp index 14e6b79..552d4c9 100644 --- a/llvm/lib/CodeGen/GlobalISel/IRTranslator.cpp +++ b/llvm/lib/CodeGen/GlobalISel/IRTranslator.cpp @@ -3984,7 +3984,7 @@ bool IRTranslator::runOnMachineFunction(MachineFunction &CurMF) { raw_string_ostream InstStr(InstStrStorage); InstStr << Inst; - R << ": '" << InstStr.str() << "'"; + R << ": '" << InstStrStorage << "'"; } reportTranslationError(*MF, *TPC, *ORE, R); diff --git a/llvm/lib/CodeGen/MIRPrinter.cpp b/llvm/lib/CodeGen/MIRPrinter.cpp index 49993f7..48c3e0d 100644 --- a/llvm/lib/CodeGen/MIRPrinter.cpp +++ b/llvm/lib/CodeGen/MIRPrinter.cpp @@ -570,7 +570,7 @@ void MIRPrinter::convertMachineMetadataNodes(yaml::MachineFunction &YMF, std::string NS; raw_string_ostream StrOS(NS); MD.second->print(StrOS, MST, MF.getFunction().getParent()); - YMF.MachineMetadataNodes.push_back(StrOS.str()); + YMF.MachineMetadataNodes.push_back(NS); } } @@ -588,7 +588,7 @@ void MIRPrinter::convert(yaml::MachineFunction &MF, yaml::MachineConstantPoolValue YamlConstant; YamlConstant.ID = ID++; - YamlConstant.Value = StrOS.str(); + YamlConstant.Value = Str; YamlConstant.Alignment = Constant.getAlign(); YamlConstant.IsTargetSpecific = Constant.isMachineConstantPoolEntry(); @@ -608,7 +608,7 @@ void MIRPrinter::convert(ModuleSlotTracker &MST, for (const auto *MBB : Table.MBBs) { raw_string_ostream StrOS(Str); StrOS << printMBBReference(*MBB); - Entry.Blocks.push_back(StrOS.str()); + Entry.Blocks.push_back(Str); Str.clear(); } YamlJTI.Entries.push_back(Entry); diff --git a/llvm/lib/CodeGen/MachineOutliner.cpp b/llvm/lib/CodeGen/MachineOutliner.cpp index 76e5952..c7ccf10 100644 --- a/llvm/lib/CodeGen/MachineOutliner.cpp +++ b/llvm/lib/CodeGen/MachineOutliner.cpp @@ -804,8 +804,7 @@ MachineFunction *MachineOutliner::createOutlinedFunction( Mg.getNameWithPrefix(MangledNameStream, F, false); DISubprogram *OutlinedSP = DB.createFunction( - Unit /* Context */, F->getName(), StringRef(MangledNameStream.str()), - Unit /* File */, + Unit /* Context */, F->getName(), StringRef(Dummy), Unit /* File */, 0 /* Line 0 is reserved for compiler-generated code. */, DB.createSubroutineType( DB.getOrCreateTypeArray(std::nullopt)), /* void type */ diff --git a/llvm/lib/CodeGen/MachineSSAUpdater.cpp b/llvm/lib/CodeGen/MachineSSAUpdater.cpp index ffe1df4..4cbb6ad 100644 --- a/llvm/lib/CodeGen/MachineSSAUpdater.cpp +++ b/llvm/lib/CodeGen/MachineSSAUpdater.cpp @@ -89,8 +89,8 @@ Register LookForIdenticalPHI(MachineBasicBlock *BB, return Register(); AvailableValsTy AVals; - for (unsigned i = 0, e = PredValues.size(); i != e; ++i) - AVals[PredValues[i].first] = PredValues[i].second; + for (const auto &[SrcBB, SrcReg] : PredValues) + AVals[SrcBB] = SrcReg; while (I != BB->end() && I->isPHI()) { bool Same = true; for (unsigned i = 1, e = I->getNumOperands(); i != e; i += 2) { @@ -196,8 +196,8 @@ Register MachineSSAUpdater::GetValueInMiddleOfBlock(MachineBasicBlock *BB, InsertNewDef(TargetOpcode::PHI, BB, Loc, RegAttrs, MRI, TII); // Fill in all the predecessors of the PHI. - for (unsigned i = 0, e = PredValues.size(); i != e; ++i) - InsertedPHI.addReg(PredValues[i].second).addMBB(PredValues[i].first); + for (const auto &[SrcBB, SrcReg] : PredValues) + InsertedPHI.addReg(SrcReg).addMBB(SrcBB); // See if the PHI node can be merged to a single value. This can happen in // loop cases when we get a PHI of itself and one other value. diff --git a/llvm/lib/CodeGen/MachineScheduler.cpp b/llvm/lib/CodeGen/MachineScheduler.cpp index cf72f74..4a6d5ed 100644 --- a/llvm/lib/CodeGen/MachineScheduler.cpp +++ b/llvm/lib/CodeGen/MachineScheduler.cpp @@ -4413,7 +4413,7 @@ struct DOTGraphTraits<ScheduleDAGMI*> : public DefaultDOTGraphTraits { SS << "SU:" << SU->NodeNum; if (DFS) SS << " I:" << DFS->getNumInstrs(SU); - return SS.str(); + return Str; } static std::string getNodeDescription(const SUnit *SU, const ScheduleDAG *G) { diff --git a/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp b/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp index c848ce4..68dece6 100644 --- a/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp +++ b/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp @@ -1206,7 +1206,7 @@ std::string ScheduleDAGInstrs::getGraphNodeLabel(const SUnit *SU) const { oss << "<exit>"; else SU->getInstr()->print(oss, /*IsStandalone=*/true); - return oss.str(); + return s; } /// Return the basic block label. It is not necessarilly unique because a block diff --git a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp index 7429763..bc16f88 100644 --- a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp +++ b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp @@ -11749,7 +11749,7 @@ SDValue SelectionDAG::getSymbolFunctionGlobalAddress(SDValue Op, raw_string_ostream ErrorFormatter(ErrorStr); ErrorFormatter << "Undefined external symbol "; ErrorFormatter << '"' << Symbol << '"'; - report_fatal_error(Twine(ErrorFormatter.str())); + report_fatal_error(Twine(ErrorStr)); } //===----------------------------------------------------------------------===// diff --git a/llvm/lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp b/llvm/lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp index 8ce6dfc..ecdbf3e 100644 --- a/llvm/lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp +++ b/llvm/lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp @@ -1778,7 +1778,7 @@ void SelectionDAGISel::SelectAllBasicBlocks(const Function &Fn) { raw_string_ostream InstStr(InstStrStorage); InstStr << *Inst; - R << ": " << InstStr.str(); + R << ": " << InstStrStorage; } reportFastISelFailure(*MF, *ORE, R, EnableFastISelAbort > 2); @@ -1827,7 +1827,7 @@ void SelectionDAGISel::SelectAllBasicBlocks(const Function &Fn) { std::string InstStrStorage; raw_string_ostream InstStr(InstStrStorage); InstStr << *Inst; - R << ": " << InstStr.str(); + R << ": " << InstStrStorage; } reportFastISelFailure(*MF, *ORE, R, ShouldAbort); @@ -4370,5 +4370,5 @@ void SelectionDAGISel::CannotYetSelect(SDNode *N) { else Msg << "unknown intrinsic #" << iid; } - report_fatal_error(Twine(Msg.str())); + report_fatal_error(Twine(msg)); } diff --git a/llvm/lib/CodeGen/SelectionDAG/SelectionDAGPrinter.cpp b/llvm/lib/CodeGen/SelectionDAG/SelectionDAGPrinter.cpp index b66eeb6..ac28f62 100644 --- a/llvm/lib/CodeGen/SelectionDAG/SelectionDAGPrinter.cpp +++ b/llvm/lib/CodeGen/SelectionDAG/SelectionDAGPrinter.cpp @@ -299,7 +299,7 @@ std::string ScheduleDAGSDNodes::getGraphNodeLabel(const SUnit *SU) const { } else { O << "CROSS RC COPY"; } - return O.str(); + return s; } void ScheduleDAGSDNodes::getCustomGraphFeatures(GraphWriter<ScheduleDAG*> &GW) const { diff --git a/llvm/lib/CodeGen/TargetInstrInfo.cpp b/llvm/lib/CodeGen/TargetInstrInfo.cpp index e01e7b3..3cd1bb2 100644 --- a/llvm/lib/CodeGen/TargetInstrInfo.cpp +++ b/llvm/lib/CodeGen/TargetInstrInfo.cpp @@ -1750,7 +1750,7 @@ std::string TargetInstrInfo::createMIROperandComment( OS << Info; } - return OS.str(); + return Flags; } int FlagIdx = MI.findInlineAsmFlagIdx(OpIdx); @@ -1784,7 +1784,7 @@ std::string TargetInstrInfo::createMIROperandComment( F.getRegMayBeFolded()) OS << " foldable"; - return OS.str(); + return Flags; } TargetInstrInfo::PipelinerLoopInfo::~PipelinerLoopInfo() = default; diff --git a/llvm/lib/MC/ELFObjectWriter.cpp b/llvm/lib/MC/ELFObjectWriter.cpp index e7e729f..bcc6dfe 100644 --- a/llvm/lib/MC/ELFObjectWriter.cpp +++ b/llvm/lib/MC/ELFObjectWriter.cpp @@ -21,7 +21,6 @@ #include "llvm/BinaryFormat/ELF.h" #include "llvm/MC/MCAsmBackend.h" #include "llvm/MC/MCAsmInfo.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCELFObjectWriter.h" @@ -567,7 +566,7 @@ void ELFWriter::writeSymbol(const MCAssembler &Asm, SymbolTableWriter &Writer, if (ESize) { int64_t Res; - if (!ESize->evaluateKnownAbsolute(Res, *Asm.getLayout())) + if (!ESize->evaluateKnownAbsolute(Res, Asm)) report_fatal_error("Size expression must be absolute."); Size = Res; } diff --git a/llvm/lib/MC/GOFFObjectWriter.cpp b/llvm/lib/MC/GOFFObjectWriter.cpp index 9c43ef1a..76f3735 100644 --- a/llvm/lib/MC/GOFFObjectWriter.cpp +++ b/llvm/lib/MC/GOFFObjectWriter.cpp @@ -11,7 +11,6 @@ //===----------------------------------------------------------------------===// #include "llvm/BinaryFormat/GOFF.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCGOFFObjectWriter.h" #include "llvm/MC/MCValue.h" diff --git a/llvm/lib/MC/MCAssembler.cpp b/llvm/lib/MC/MCAssembler.cpp index 53bebdc..14790f5 100644 --- a/llvm/lib/MC/MCAssembler.cpp +++ b/llvm/lib/MC/MCAssembler.cpp @@ -15,7 +15,6 @@ #include "llvm/ADT/Twine.h" #include "llvm/MC/MCAsmBackend.h" #include "llvm/MC/MCAsmInfo.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCCodeEmitter.h" #include "llvm/MC/MCCodeView.h" #include "llvm/MC/MCContext.h" @@ -187,7 +186,7 @@ bool MCAssembler::evaluateFixup(const MCFixup &Fixup, const MCFragment *DF, MCContext &Ctx = getContext(); Value = 0; WasForced = false; - if (!Expr->evaluateAsRelocatable(Target, Layout, &Fixup)) { + if (!Expr->evaluateAsRelocatable(Target, this, &Fixup)) { Ctx.reportError(Fixup.getLoc(), "expected relocatable expression"); return true; } @@ -288,7 +287,7 @@ uint64_t MCAssembler::computeFragmentSize(const MCFragment &F) const { case MCFragment::FT_Fill: { auto &FF = cast<MCFillFragment>(F); int64_t NumValues = 0; - if (!FF.getNumValues().evaluateKnownAbsolute(NumValues, *Layout)) { + if (!FF.getNumValues().evaluateKnownAbsolute(NumValues, *this)) { getContext().reportError(FF.getLoc(), "expected assembly-time absolute expression"); return 0; @@ -381,8 +380,6 @@ uint64_t MCAssembler::computeFragmentSize(const MCFragment &F) const { llvm_unreachable("invalid fragment kind"); } -MCAsmLayout::MCAsmLayout(MCAssembler &Asm) : Assembler(Asm) {} - // Compute the amount of padding required before the fragment \p F to // obey bundling restrictions, where \p FOffset is the fragment's offset in // its section and \p FSize is the fragment's size. @@ -547,7 +544,7 @@ uint64_t MCAssembler::getSymbolOffset(const MCSymbol &S) const { } const MCSymbol *MCAssembler::getBaseSymbol(const MCSymbol &Symbol) const { - assert(Layout); + assert(HasLayout); if (!Symbol.isVariable()) return &Symbol; @@ -584,6 +581,7 @@ const MCSymbol *MCAssembler::getBaseSymbol(const MCSymbol &Symbol) const { } uint64_t MCAssembler::getSectionAddressSize(const MCSection &Sec) const { + assert(HasLayout); // The size is the last fragment's end offset. const MCFragment &F = *Sec.curFragList()->Tail; return getFragmentOffset(F) + computeFragmentSize(F); @@ -937,7 +935,7 @@ MCAssembler::handleFixup(MCFragment &F, const MCFixup &Fixup, return std::make_tuple(Target, FixedValue, IsResolved); } -void MCAssembler::layout(MCAsmLayout &Layout) { +void MCAssembler::layout() { assert(getBackendPtr() && "Expected assembler backend"); DEBUG_WITH_TYPE("mc-dump", { errs() << "assembler backend - pre-layout\n--\n"; @@ -968,7 +966,7 @@ void MCAssembler::layout(MCAsmLayout &Layout) { } // Layout until everything fits. - this->Layout = &Layout; + this->HasLayout = true; while (layoutOnce()) { if (getContext().hadError()) return; @@ -984,7 +982,7 @@ void MCAssembler::layout(MCAsmLayout &Layout) { dump(); }); // Finalize the layout, including fragment lowering. - finishLayout(Layout); + getBackend().finishLayout(*this); DEBUG_WITH_TYPE("mc-dump", { errs() << "assembler backend - final-layout\n--\n"; @@ -1074,14 +1072,12 @@ void MCAssembler::layout(MCAsmLayout &Layout) { } void MCAssembler::Finish() { - // Create the layout object. - MCAsmLayout Layout(*this); - layout(Layout); + layout(); // Write the object file. stats::ObjectBytes += getWriter().writeObject(*this); - this->Layout = nullptr; + HasLayout = false; } bool MCAssembler::fixupNeedsRelaxation(const MCFixup &Fixup, @@ -1150,8 +1146,8 @@ bool MCAssembler::relaxLEB(MCLEBFragment &LF) { // requires that .uleb128 A-B is foldable where A and B reside in different // fragments. This is used by __gcc_except_table. bool Abs = getSubsectionsViaSymbols() - ? LF.getValue().evaluateKnownAbsolute(Value, *Layout) - : LF.getValue().evaluateAsAbsolute(Value, *Layout); + ? LF.getValue().evaluateKnownAbsolute(Value, *this) + : LF.getValue().evaluateAsAbsolute(Value, *this); if (!Abs) { bool Relaxed, UseZeroPad; std::tie(Relaxed, UseZeroPad) = getBackend().relaxLEB128(*this, LF, Value); @@ -1248,7 +1244,7 @@ bool MCAssembler::relaxDwarfLineAddr(MCDwarfLineAddrFragment &DF) { MCContext &Context = getContext(); uint64_t OldSize = DF.getContents().size(); int64_t AddrDelta; - bool Abs = DF.getAddrDelta().evaluateKnownAbsolute(AddrDelta, *Layout); + bool Abs = DF.getAddrDelta().evaluateKnownAbsolute(AddrDelta, *this); assert(Abs && "We created a line delta with an invalid expression"); (void)Abs; int64_t LineDelta; @@ -1269,7 +1265,7 @@ bool MCAssembler::relaxDwarfCallFrameFragment(MCDwarfCallFrameFragment &DF) { MCContext &Context = getContext(); int64_t Value; - bool Abs = DF.getAddrDelta().evaluateAsAbsolute(Value, *Layout); + bool Abs = DF.getAddrDelta().evaluateAsAbsolute(Value, *this); if (!Abs) { getContext().reportError(DF.getAddrDelta().getLoc(), "invalid CFI advance_loc expression"); @@ -1288,20 +1284,20 @@ bool MCAssembler::relaxDwarfCallFrameFragment(MCDwarfCallFrameFragment &DF) { bool MCAssembler::relaxCVInlineLineTable(MCCVInlineLineTableFragment &F) { unsigned OldSize = F.getContents().size(); - getContext().getCVContext().encodeInlineLineTable(*Layout, F); + getContext().getCVContext().encodeInlineLineTable(*this, F); return OldSize != F.getContents().size(); } bool MCAssembler::relaxCVDefRange(MCCVDefRangeFragment &F) { unsigned OldSize = F.getContents().size(); - getContext().getCVContext().encodeDefRange(*Layout, F); + getContext().getCVContext().encodeDefRange(*this, F); return OldSize != F.getContents().size(); } bool MCAssembler::relaxPseudoProbeAddr(MCPseudoProbeAddrFragment &PF) { uint64_t OldSize = PF.getContents().size(); int64_t AddrDelta; - bool Abs = PF.getAddrDelta().evaluateKnownAbsolute(AddrDelta, *Layout); + bool Abs = PF.getAddrDelta().evaluateKnownAbsolute(AddrDelta, *this); assert(Abs && "We created a pseudo probe with an invalid expression"); (void)Abs; SmallVectorImpl<char> &Data = PF.getContents(); @@ -1350,11 +1346,6 @@ bool MCAssembler::layoutOnce() { return Changed; } -void MCAssembler::finishLayout(MCAsmLayout &Layout) { - assert(getBackendPtr() && "Expected assembler backend"); - getBackend().finishLayout(*this, Layout); -} - #if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP) LLVM_DUMP_METHOD void MCAssembler::dump() const{ raw_ostream &OS = errs(); diff --git a/llvm/lib/MC/MCCodeView.cpp b/llvm/lib/MC/MCCodeView.cpp index 89b28b4..792a132 100644 --- a/llvm/lib/MC/MCCodeView.cpp +++ b/llvm/lib/MC/MCCodeView.cpp @@ -16,7 +16,6 @@ #include "llvm/DebugInfo/CodeView/CodeView.h" #include "llvm/DebugInfo/CodeView/Line.h" #include "llvm/DebugInfo/CodeView/SymbolRecord.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCObjectStreamer.h" @@ -465,16 +464,16 @@ MCFragment *CodeViewContext::emitDefRange( return F; } -static unsigned computeLabelDiff(MCAsmLayout &Layout, const MCSymbol *Begin, +static unsigned computeLabelDiff(const MCAssembler &Asm, const MCSymbol *Begin, const MCSymbol *End) { - MCContext &Ctx = Layout.getAssembler().getContext(); + MCContext &Ctx = Asm.getContext(); MCSymbolRefExpr::VariantKind Variant = MCSymbolRefExpr::VK_None; const MCExpr *BeginRef = MCSymbolRefExpr::create(Begin, Variant, Ctx), *EndRef = MCSymbolRefExpr::create(End, Variant, Ctx); const MCExpr *AddrDelta = MCBinaryExpr::create(MCBinaryExpr::Sub, EndRef, BeginRef, Ctx); int64_t Result; - bool Success = AddrDelta->evaluateKnownAbsolute(Result, Layout); + bool Success = AddrDelta->evaluateKnownAbsolute(Result, Asm); assert(Success && "failed to evaluate label difference as absolute"); (void)Success; assert(Result >= 0 && "negative label difference requested"); @@ -482,7 +481,7 @@ static unsigned computeLabelDiff(MCAsmLayout &Layout, const MCSymbol *Begin, return unsigned(Result); } -void CodeViewContext::encodeInlineLineTable(MCAsmLayout &Layout, +void CodeViewContext::encodeInlineLineTable(const MCAssembler &Asm, MCCVInlineLineTableFragment &Frag) { size_t LocBegin; size_t LocEnd; @@ -549,7 +548,7 @@ void CodeViewContext::encodeInlineLineTable(MCAsmLayout &Layout, // We've hit a cv_loc not attributed to this inline call site. Use this // label to end the PC range. if (HaveOpenRange) { - unsigned Length = computeLabelDiff(Layout, LastLabel, Loc.getLabel()); + unsigned Length = computeLabelDiff(Asm, LastLabel, Loc.getLabel()); compressAnnotation(BinaryAnnotationsOpCode::ChangeCodeLength, Buffer); compressAnnotation(Length, Buffer); LastLabel = Loc.getLabel(); @@ -579,7 +578,7 @@ void CodeViewContext::encodeInlineLineTable(MCAsmLayout &Layout, int LineDelta = CurSourceLoc.Line - LastSourceLoc.Line; unsigned EncodedLineDelta = encodeSignedNumber(LineDelta); - unsigned CodeDelta = computeLabelDiff(Layout, LastLabel, Loc.getLabel()); + unsigned CodeDelta = computeLabelDiff(Asm, LastLabel, Loc.getLabel()); if (EncodedLineDelta < 0x8 && CodeDelta <= 0xf) { // The ChangeCodeOffsetAndLineOffset combination opcode is used when the // encoded line delta uses 3 or fewer set bits and the code offset fits @@ -605,23 +604,23 @@ void CodeViewContext::encodeInlineLineTable(MCAsmLayout &Layout, assert(HaveOpenRange); unsigned EndSymLength = - computeLabelDiff(Layout, LastLabel, Frag.getFnEndSym()); + computeLabelDiff(Asm, LastLabel, Frag.getFnEndSym()); unsigned LocAfterLength = ~0U; ArrayRef<MCCVLoc> LocAfter = getLinesForExtent(LocEnd, LocEnd + 1); if (!LocAfter.empty()) { // Only try to compute this difference if we're in the same section. const MCCVLoc &Loc = LocAfter[0]; if (&Loc.getLabel()->getSection() == &LastLabel->getSection()) - LocAfterLength = computeLabelDiff(Layout, LastLabel, Loc.getLabel()); + LocAfterLength = computeLabelDiff(Asm, LastLabel, Loc.getLabel()); } compressAnnotation(BinaryAnnotationsOpCode::ChangeCodeLength, Buffer); compressAnnotation(std::min(EndSymLength, LocAfterLength), Buffer); } -void CodeViewContext::encodeDefRange(MCAsmLayout &Layout, +void CodeViewContext::encodeDefRange(const MCAssembler &Asm, MCCVDefRangeFragment &Frag) { - MCContext &Ctx = Layout.getAssembler().getContext(); + MCContext &Ctx = Asm.getContext(); SmallVectorImpl<char> &Contents = Frag.getContents(); Contents.clear(); SmallVectorImpl<MCFixup> &Fixups = Frag.getFixups(); @@ -633,8 +632,8 @@ void CodeViewContext::encodeDefRange(MCAsmLayout &Layout, const MCSymbol *LastLabel = nullptr; for (std::pair<const MCSymbol *, const MCSymbol *> Range : Frag.getRanges()) { unsigned GapSize = - LastLabel ? computeLabelDiff(Layout, LastLabel, Range.first) : 0; - unsigned RangeSize = computeLabelDiff(Layout, Range.first, Range.second); + LastLabel ? computeLabelDiff(Asm, LastLabel, Range.first) : 0; + unsigned RangeSize = computeLabelDiff(Asm, Range.first, Range.second); GapAndRangeSizes.push_back({GapSize, RangeSize}); LastLabel = Range.second; } diff --git a/llvm/lib/MC/MCDXContainerWriter.cpp b/llvm/lib/MC/MCDXContainerWriter.cpp index 002fe2f..6b42d38 100644 --- a/llvm/lib/MC/MCDXContainerWriter.cpp +++ b/llvm/lib/MC/MCDXContainerWriter.cpp @@ -8,7 +8,6 @@ #include "llvm/MC/MCDXContainerWriter.h" #include "llvm/BinaryFormat/DXContainer.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCSection.h" diff --git a/llvm/lib/MC/MCExpr.cpp b/llvm/lib/MC/MCExpr.cpp index af9dcff..b42a668 100644 --- a/llvm/lib/MC/MCExpr.cpp +++ b/llvm/lib/MC/MCExpr.cpp @@ -12,7 +12,6 @@ #include "llvm/Config/llvm-config.h" #include "llvm/MC/MCAsmBackend.h" #include "llvm/MC/MCAsmInfo.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCObjectWriter.h" @@ -545,37 +544,29 @@ void MCTargetExpr::anchor() {} /* *** */ bool MCExpr::evaluateAsAbsolute(int64_t &Res) const { - return evaluateAsAbsolute(Res, nullptr, nullptr, nullptr, false); -} - -bool MCExpr::evaluateAsAbsolute(int64_t &Res, - const MCAsmLayout &Layout) const { - return evaluateAsAbsolute(Res, &Layout.getAssembler(), &Layout, nullptr, false); + return evaluateAsAbsolute(Res, nullptr, nullptr, false); } bool MCExpr::evaluateAsAbsolute(int64_t &Res, const MCAssembler &Asm, const SectionAddrMap &Addrs) const { // Setting InSet causes us to absolutize differences across sections and that // is what the MachO writer uses Addrs for. - return evaluateAsAbsolute(Res, &Asm, Asm.getLayout(), &Addrs, true); + return evaluateAsAbsolute(Res, &Asm, &Addrs, true); } bool MCExpr::evaluateAsAbsolute(int64_t &Res, const MCAssembler &Asm) const { - return evaluateAsAbsolute(Res, &Asm, nullptr, nullptr, false); + return evaluateAsAbsolute(Res, &Asm, nullptr, false); } bool MCExpr::evaluateAsAbsolute(int64_t &Res, const MCAssembler *Asm) const { - return evaluateAsAbsolute(Res, Asm, nullptr, nullptr, false); + return evaluateAsAbsolute(Res, Asm, nullptr, false); } -bool MCExpr::evaluateKnownAbsolute(int64_t &Res, - const MCAsmLayout &Layout) const { - return evaluateAsAbsolute(Res, &Layout.getAssembler(), &Layout, nullptr, - true); +bool MCExpr::evaluateKnownAbsolute(int64_t &Res, const MCAssembler &Asm) const { + return evaluateAsAbsolute(Res, &Asm, nullptr, true); } bool MCExpr::evaluateAsAbsolute(int64_t &Res, const MCAssembler *Asm, - const MCAsmLayout *Layout, const SectionAddrMap *Addrs, bool InSet) const { MCValue Value; @@ -634,7 +625,7 @@ static void AttemptToFoldSymbolOffsetDifference( // separated by a linker-relaxable instruction. If the section contains // instructions and InSet is false (not expressions in directive like // .size/.fill), disable the fast path. - const MCAsmLayout *Layout = Asm->getLayout(); + bool Layout = Asm->hasLayout(); if (Layout && (InSet || !SecA.hasInstructions() || !(Asm->getContext().getTargetTriple().isRISCV() || Asm->getContext().getTargetTriple().isLoongArch()))) { @@ -795,11 +786,9 @@ static bool evaluateSymbolicAdd(const MCAssembler *Asm, return true; } -bool MCExpr::evaluateAsRelocatable(MCValue &Res, - const MCAsmLayout *Layout, +bool MCExpr::evaluateAsRelocatable(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const { - MCAssembler *Assembler = Layout ? &Layout->getAssembler() : nullptr; - return evaluateAsRelocatableImpl(Res, Assembler, Fixup, nullptr, false); + return evaluateAsRelocatableImpl(Res, Asm, Fixup, nullptr, false); } bool MCExpr::evaluateAsValue(MCValue &Res, const MCAssembler &Asm) const { @@ -827,11 +816,9 @@ bool MCExpr::evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const SectionAddrMap *Addrs, bool InSet) const { ++stats::MCExprEvaluate; - MCAsmLayout *Layout = Asm ? Asm->getLayout() : nullptr; switch (getKind()) { case Target: - return cast<MCTargetExpr>(this)->evaluateAsRelocatableImpl(Res, Layout, - Fixup); + return cast<MCTargetExpr>(this)->evaluateAsRelocatableImpl(Res, Asm, Fixup); case Constant: Res = MCValue::get(cast<MCConstantExpr>(this)->getValue()); @@ -841,6 +828,7 @@ bool MCExpr::evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCSymbolRefExpr *SRE = cast<MCSymbolRefExpr>(this); const MCSymbol &Sym = SRE->getSymbol(); const auto Kind = SRE->getKind(); + bool Layout = Asm && Asm->hasLayout(); // Evaluate recursively if this is a variable. if (Sym.isVariable() && (Kind == MCSymbolRefExpr::VK_None || Layout) && diff --git a/llvm/lib/MC/MCLinkerOptimizationHint.cpp b/llvm/lib/MC/MCLinkerOptimizationHint.cpp index 9ab3218..a0901af 100644 --- a/llvm/lib/MC/MCLinkerOptimizationHint.cpp +++ b/llvm/lib/MC/MCLinkerOptimizationHint.cpp @@ -7,7 +7,6 @@ //===----------------------------------------------------------------------===// #include "llvm/MC/MCLinkerOptimizationHint.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCMachObjectWriter.h" #include "llvm/Support/LEB128.h" #include "llvm/Support/raw_ostream.h" @@ -24,23 +23,24 @@ using namespace llvm; // - Its argN. // <arg1> to <argN> are absolute addresses in the object file, i.e., // relative addresses from the beginning of the object file. -void MCLOHDirective::emit_impl(raw_ostream &OutStream, - const MachObjectWriter &ObjWriter, - const MCAsmLayout &Layout) const { +void MCLOHDirective::emit_impl(const MCAssembler &Asm, raw_ostream &OutStream, + const MachObjectWriter &ObjWriter + +) const { encodeULEB128(Kind, OutStream); encodeULEB128(Args.size(), OutStream); for (const MCSymbol *Arg : Args) - encodeULEB128(ObjWriter.getSymbolAddress(*Arg, Layout), OutStream); + encodeULEB128(ObjWriter.getSymbolAddress(*Arg, Asm), OutStream); } -void MCLOHDirective::emit(MachObjectWriter &ObjWriter, - const MCAsmLayout &Layout) const { +void MCLOHDirective::emit(const MCAssembler &Asm, + MachObjectWriter &ObjWriter) const { raw_ostream &OutStream = ObjWriter.W.OS; - emit_impl(OutStream, ObjWriter, Layout); + emit_impl(Asm, OutStream, ObjWriter); } -uint64_t MCLOHDirective::getEmitSize(const MachObjectWriter &ObjWriter, - const MCAsmLayout &Layout) const { +uint64_t MCLOHDirective::getEmitSize(const MCAssembler &Asm, + const MachObjectWriter &ObjWriter) const { class raw_counting_ostream : public raw_ostream { uint64_t Count = 0; @@ -54,6 +54,6 @@ uint64_t MCLOHDirective::getEmitSize(const MachObjectWriter &ObjWriter, }; raw_counting_ostream OutStream; - emit_impl(OutStream, ObjWriter, Layout); + emit_impl(Asm, OutStream, ObjWriter); return OutStream.tell(); } diff --git a/llvm/lib/MC/MachObjectWriter.cpp b/llvm/lib/MC/MachObjectWriter.cpp index 258dd69..3f5bdd8 100644 --- a/llvm/lib/MC/MachObjectWriter.cpp +++ b/llvm/lib/MC/MachObjectWriter.cpp @@ -11,7 +11,6 @@ #include "llvm/ADT/iterator_range.h" #include "llvm/BinaryFormat/MachO.h" #include "llvm/MC/MCAsmBackend.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAsmInfoDarwin.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" @@ -91,7 +90,7 @@ MachObjectWriter::getFragmentAddress(const MCAssembler &Asm, } uint64_t MachObjectWriter::getSymbolAddress(const MCSymbol &S, - const MCAsmLayout &Layout) const { + const MCAssembler &Asm) const { // If this is a variable, then recursively evaluate now. if (S.isVariable()) { if (const MCConstantExpr *C = @@ -99,7 +98,7 @@ uint64_t MachObjectWriter::getSymbolAddress(const MCSymbol &S, return C->getValue(); MCValue Target; - if (!S.getVariableValue()->evaluateAsRelocatable(Target, &Layout, nullptr)) + if (!S.getVariableValue()->evaluateAsRelocatable(Target, &Asm, nullptr)) report_fatal_error("unable to evaluate offset for variable '" + S.getName() + "'"); @@ -113,14 +112,14 @@ uint64_t MachObjectWriter::getSymbolAddress(const MCSymbol &S, uint64_t Address = Target.getConstant(); if (Target.getSymA()) - Address += getSymbolAddress(Target.getSymA()->getSymbol(), Layout); + Address += getSymbolAddress(Target.getSymA()->getSymbol(), Asm); if (Target.getSymB()) - Address += getSymbolAddress(Target.getSymB()->getSymbol(), Layout); + Address += getSymbolAddress(Target.getSymB()->getSymbol(), Asm); return Address; } return getSectionAddress(S.getFragment()->getParent()) + - Layout.getAssembler().getSymbolOffset(S); + Asm.getSymbolOffset(S); } uint64_t MachObjectWriter::getPaddingSize(const MCAssembler &Asm, @@ -370,8 +369,7 @@ const MCSymbol &MachObjectWriter::findAliasedSymbol(const MCSymbol &Sym) const { return *S; } -void MachObjectWriter::writeNlist(MachSymbolData &MSD, - const MCAsmLayout &Layout) { +void MachObjectWriter::writeNlist(MachSymbolData &MSD, const MCAssembler &Asm) { const MCSymbol *Symbol = MSD.Symbol; const MCSymbol &Data = *Symbol; const MCSymbol *AliasedSymbol = &findAliasedSymbol(*Symbol); @@ -415,7 +413,7 @@ void MachObjectWriter::writeNlist(MachSymbolData &MSD, if (IsAlias && Symbol->isUndefined()) Address = AliaseeInfo->StringIndex; else if (Symbol->isDefined()) - Address = getSymbolAddress(OrigSymbol, Layout); + Address = getSymbolAddress(OrigSymbol, Asm); else if (Symbol->isCommon()) { // Common symbols are encoded with the size in the address // field, and their alignment in the flags. @@ -780,7 +778,6 @@ void MachObjectWriter::populateAddrSigSection(MCAssembler &Asm) { } uint64_t MachObjectWriter::writeObject(MCAssembler &Asm) { - auto &Layout = *Asm.getLayout(); uint64_t StartOffset = W.OS.tell(); populateAddrSigSection(Asm); @@ -842,7 +839,7 @@ uint64_t MachObjectWriter::writeObject(MCAssembler &Asm) { } // Add the loh load command size, if used. - uint64_t LOHRawSize = Asm.getLOHContainer().getEmitSize(*this, Layout); + uint64_t LOHRawSize = Asm.getLOHContainer().getEmitSize(Asm, *this); uint64_t LOHSize = alignTo(LOHRawSize, is64Bit() ? 8 : 4); if (LOHSize) { ++NumLoadCommands; @@ -1036,10 +1033,10 @@ uint64_t MachObjectWriter::writeObject(MCAssembler &Asm) { it = Asm.data_region_begin(), ie = Asm.data_region_end(); it != ie; ++it) { const DataRegionData *Data = &(*it); - uint64_t Start = getSymbolAddress(*Data->Start, Layout); + uint64_t Start = getSymbolAddress(*Data->Start, Asm); uint64_t End; if (Data->End) - End = getSymbolAddress(*Data->End, Layout); + End = getSymbolAddress(*Data->End, Asm); else report_fatal_error("Data region not terminated"); @@ -1058,7 +1055,7 @@ uint64_t MachObjectWriter::writeObject(MCAssembler &Asm) { #ifndef NDEBUG unsigned Start = W.OS.tell(); #endif - Asm.getLOHContainer().emit(*this, Layout); + Asm.getLOHContainer().emit(Asm, *this); // Pad to a multiple of the pointer size. W.OS.write_zeros( offsetToAlignment(LOHRawSize, is64Bit() ? Align(8) : Align(4))); @@ -1095,7 +1092,7 @@ uint64_t MachObjectWriter::writeObject(MCAssembler &Asm) { for (auto *SymbolData : {&LocalSymbolData, &ExternalSymbolData, &UndefinedSymbolData}) for (MachSymbolData &Entry : *SymbolData) - writeNlist(Entry, Layout); + writeNlist(Entry, Asm); // Write the string table. StringTable.write(W.OS); diff --git a/llvm/lib/MC/WasmObjectWriter.cpp b/llvm/lib/MC/WasmObjectWriter.cpp index 5602c9b..f25dc92 100644 --- a/llvm/lib/MC/WasmObjectWriter.cpp +++ b/llvm/lib/MC/WasmObjectWriter.cpp @@ -15,7 +15,6 @@ #include "llvm/BinaryFormat/WasmTraits.h" #include "llvm/Config/llvm-config.h" #include "llvm/MC/MCAsmBackend.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCExpr.h" @@ -299,11 +298,10 @@ private: void executePostLayoutBinding(MCAssembler &Asm) override; void prepareImports(SmallVectorImpl<wasm::WasmImport> &Imports, - MCAssembler &Asm, const MCAsmLayout &Layout); + MCAssembler &Asm); uint64_t writeObject(MCAssembler &Asm) override; - uint64_t writeOneObject(MCAssembler &Asm, const MCAsmLayout &Layout, - DwoMode Mode); + uint64_t writeOneObject(MCAssembler &Asm, DwoMode Mode); void writeString(const StringRef Str) { encodeULEB128(Str.size(), W->OS); @@ -334,9 +332,9 @@ private: void writeElemSection(const MCSymbolWasm *IndirectFunctionTable, ArrayRef<uint32_t> TableElems); void writeDataCountSection(); - uint32_t writeCodeSection(const MCAssembler &Asm, const MCAsmLayout &Layout, + uint32_t writeCodeSection(const MCAssembler &Asm, ArrayRef<WasmFunction> Functions); - uint32_t writeDataSection(const MCAsmLayout &Layout); + uint32_t writeDataSection(const MCAssembler &Asm); void writeTagSection(ArrayRef<uint32_t> TagTypes); void writeGlobalSection(ArrayRef<wasm::WasmGlobal> Globals); void writeTableSection(ArrayRef<wasm::WasmTable> Tables); @@ -347,13 +345,13 @@ private: ArrayRef<std::pair<uint16_t, uint32_t>> InitFuncs, const std::map<StringRef, std::vector<WasmComdatEntry>> &Comdats); void writeCustomSection(WasmCustomSection &CustomSection, - const MCAssembler &Asm, const MCAsmLayout &Layout); + const MCAssembler &Asm); void writeCustomRelocSections(); uint64_t getProvisionalValue(const MCAssembler &Asm, const WasmRelocationEntry &RelEntry); void applyRelocations(ArrayRef<WasmRelocationEntry> Relocations, - uint64_t ContentsOffset, const MCAsmLayout &Layout); + uint64_t ContentsOffset, const MCAssembler &Asm); uint32_t getRelocationIndexValue(const WasmRelocationEntry &RelEntry); uint32_t getFunctionType(const MCSymbolWasm &Symbol); @@ -764,7 +762,7 @@ WasmObjectWriter::getRelocationIndexValue(const WasmRelocationEntry &RelEntry) { // directly. void WasmObjectWriter::applyRelocations( ArrayRef<WasmRelocationEntry> Relocations, uint64_t ContentsOffset, - const MCAsmLayout &Layout) { + const MCAssembler &Asm) { auto &Stream = static_cast<raw_pwrite_stream &>(W->OS); for (const WasmRelocationEntry &RelEntry : Relocations) { uint64_t Offset = ContentsOffset + @@ -772,7 +770,7 @@ void WasmObjectWriter::applyRelocations( RelEntry.Offset; LLVM_DEBUG(dbgs() << "applyRelocation: " << RelEntry << "\n"); - uint64_t Value = getProvisionalValue(Layout.getAssembler(), RelEntry); + uint64_t Value = getProvisionalValue(Asm, RelEntry); switch (RelEntry.Type) { case wasm::R_WASM_FUNCTION_INDEX_LEB: @@ -1049,7 +1047,6 @@ void WasmObjectWriter::writeDataCountSection() { } uint32_t WasmObjectWriter::writeCodeSection(const MCAssembler &Asm, - const MCAsmLayout &Layout, ArrayRef<WasmFunction> Functions) { if (Functions.empty()) return 0; @@ -1069,13 +1066,13 @@ uint32_t WasmObjectWriter::writeCodeSection(const MCAssembler &Asm, } // Apply fixups. - applyRelocations(CodeRelocations, Section.ContentsOffset, Layout); + applyRelocations(CodeRelocations, Section.ContentsOffset, Asm); endSection(Section); return Section.Index; } -uint32_t WasmObjectWriter::writeDataSection(const MCAsmLayout &Layout) { +uint32_t WasmObjectWriter::writeDataSection(const MCAssembler &Asm) { if (DataSegments.empty()) return 0; @@ -1100,7 +1097,7 @@ uint32_t WasmObjectWriter::writeDataSection(const MCAsmLayout &Layout) { } // Apply fixups. - applyRelocations(DataRelocations, Section.ContentsOffset, Layout); + applyRelocations(DataRelocations, Section.ContentsOffset, Asm); endSection(Section); return Section.Index; @@ -1239,8 +1236,7 @@ void WasmObjectWriter::writeLinkingMetaDataSection( } void WasmObjectWriter::writeCustomSection(WasmCustomSection &CustomSection, - const MCAssembler &Asm, - const MCAsmLayout &Layout) { + const MCAssembler &Asm) { SectionBookkeeping Section; auto *Sec = CustomSection.Section; startCustomSection(Section, CustomSection.Name); @@ -1255,7 +1251,7 @@ void WasmObjectWriter::writeCustomSection(WasmCustomSection &CustomSection, // Apply fixups. auto &Relocations = CustomSectionsRelocations[CustomSection.Section]; - applyRelocations(Relocations, CustomSection.OutputContentsOffset, Layout); + applyRelocations(Relocations, CustomSection.OutputContentsOffset, Asm); } uint32_t WasmObjectWriter::getFunctionType(const MCSymbolWasm &Symbol) { @@ -1331,8 +1327,7 @@ static bool isInSymtab(const MCSymbolWasm &Sym) { } void WasmObjectWriter::prepareImports( - SmallVectorImpl<wasm::WasmImport> &Imports, MCAssembler &Asm, - const MCAsmLayout &Layout) { + SmallVectorImpl<wasm::WasmImport> &Imports, MCAssembler &Asm) { // For now, always emit the memory import, since loads and stores are not // valid without it. In the future, we could perhaps be more clever and omit // it if there are no loads or stores. @@ -1437,22 +1432,20 @@ void WasmObjectWriter::prepareImports( } uint64_t WasmObjectWriter::writeObject(MCAssembler &Asm) { - auto &Layout = *Asm.getLayout(); support::endian::Writer MainWriter(*OS, llvm::endianness::little); W = &MainWriter; if (IsSplitDwarf) { - uint64_t TotalSize = writeOneObject(Asm, Layout, DwoMode::NonDwoOnly); + uint64_t TotalSize = writeOneObject(Asm, DwoMode::NonDwoOnly); assert(DwoOS); support::endian::Writer DwoWriter(*DwoOS, llvm::endianness::little); W = &DwoWriter; - return TotalSize + writeOneObject(Asm, Layout, DwoMode::DwoOnly); + return TotalSize + writeOneObject(Asm, DwoMode::DwoOnly); } else { - return writeOneObject(Asm, Layout, DwoMode::AllSections); + return writeOneObject(Asm, DwoMode::AllSections); } } uint64_t WasmObjectWriter::writeOneObject(MCAssembler &Asm, - const MCAsmLayout &Layout, DwoMode Mode) { uint64_t StartOffset = W->OS.tell(); SectionCount = 0; @@ -1472,9 +1465,8 @@ uint64_t WasmObjectWriter::writeOneObject(MCAssembler &Asm, SmallVector<std::pair<uint16_t, uint32_t>, 2> InitFuncs; std::map<StringRef, std::vector<WasmComdatEntry>> Comdats; uint64_t DataSize = 0; - if (Mode != DwoMode::DwoOnly) { - prepareImports(Imports, Asm, Layout); - } + if (Mode != DwoMode::DwoOnly) + prepareImports(Imports, Asm); // Populate DataSegments and CustomSections, which must be done before // populating DataLocations. @@ -1633,7 +1625,7 @@ uint64_t WasmObjectWriter::writeOneObject(MCAssembler &Asm, WS.getName()); int64_t Size = 0; - if (!WS.getSize()->evaluateAsAbsolute(Size, Layout)) + if (!WS.getSize()->evaluateAsAbsolute(Size, Asm)) report_fatal_error(".size expression must be evaluatable"); auto &DataSection = static_cast<MCSectionWasm &>(WS.getSection()); @@ -1752,7 +1744,7 @@ uint64_t WasmObjectWriter::writeOneObject(MCAssembler &Asm, // size of the alias. When an offset from the base is involved this // can result in a offset + size goes past the end of the data section // which out object format doesn't support. So we must clamp it. - if (!Base->getSize()->evaluateAsAbsolute(Size, Layout)) + if (!Base->getSize()->evaluateAsAbsolute(Size, Asm)) report_fatal_error(".size expression must be evaluatable"); const WasmDataSegment &Segment = DataSegments[DataSection.getSegmentIndex()]; @@ -1926,8 +1918,8 @@ uint64_t WasmObjectWriter::writeOneObject(MCAssembler &Asm, TableElems); writeDataCountSection(); - CodeSectionIndex = writeCodeSection(Asm, Layout, Functions); - DataSectionIndex = writeDataSection(Layout); + CodeSectionIndex = writeCodeSection(Asm, Functions); + DataSectionIndex = writeDataSection(Asm); } // The Sections in the COMDAT list have placeholder indices (their index among @@ -1940,7 +1932,7 @@ uint64_t WasmObjectWriter::writeOneObject(MCAssembler &Asm, } } for (auto &CustomSection : CustomSections) - writeCustomSection(CustomSection, Asm, Layout); + writeCustomSection(CustomSection, Asm); if (Mode != DwoMode::DwoOnly) { writeLinkingMetaDataSection(SymbolInfos, InitFuncs, Comdats); @@ -1950,9 +1942,9 @@ uint64_t WasmObjectWriter::writeOneObject(MCAssembler &Asm, } writeCustomRelocSections(); if (ProducersSection) - writeCustomSection(*ProducersSection, Asm, Layout); + writeCustomSection(*ProducersSection, Asm); if (TargetFeaturesSection) - writeCustomSection(*TargetFeaturesSection, Asm, Layout); + writeCustomSection(*TargetFeaturesSection, Asm); // TODO: Translate the .comment section to the output. return W->OS.tell() - StartOffset; diff --git a/llvm/lib/MC/WinCOFFObjectWriter.cpp b/llvm/lib/MC/WinCOFFObjectWriter.cpp index cd20985..c0bad19 100644 --- a/llvm/lib/MC/WinCOFFObjectWriter.cpp +++ b/llvm/lib/MC/WinCOFFObjectWriter.cpp @@ -18,7 +18,6 @@ #include "llvm/ADT/StringRef.h" #include "llvm/ADT/Twine.h" #include "llvm/BinaryFormat/COFF.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCExpr.h" @@ -188,15 +187,13 @@ private: void WriteAuxiliarySymbols(const COFFSymbol::AuxiliarySymbols &S); void writeSectionHeaders(); void WriteRelocation(const COFF::relocation &R); - uint32_t writeSectionContents(MCAssembler &Asm, const MCAsmLayout &Layout, - const MCSection &MCSec); - void writeSection(MCAssembler &Asm, const MCAsmLayout &Layout, - const COFFSection &Sec); + uint32_t writeSectionContents(MCAssembler &Asm, const MCSection &MCSec); + void writeSection(MCAssembler &Asm, const COFFSection &Sec); void createFileSymbols(MCAssembler &Asm); void setWeakDefaultNames(); void assignSectionNumbers(); - void assignFileOffsets(MCAssembler &Asm, const MCAsmLayout &Layout); + void assignFileOffsets(MCAssembler &Asm); }; class WinCOFFObjectWriter : public MCObjectWriter { @@ -601,7 +598,6 @@ void WinCOFFWriter::WriteRelocation(const COFF::relocation &R) { // "Asm.writeSectionData(&MCSec)", but it's a bit complicated // because it needs to compute a CRC. uint32_t WinCOFFWriter::writeSectionContents(MCAssembler &Asm, - const MCAsmLayout &Layout, const MCSection &MCSec) { // Save the contents of the section to a temporary buffer, we need this // to CRC the data before we dump it into the object file. @@ -619,8 +615,7 @@ uint32_t WinCOFFWriter::writeSectionContents(MCAssembler &Asm, return JC.getCRC(); } -void WinCOFFWriter::writeSection(MCAssembler &Asm, const MCAsmLayout &Layout, - const COFFSection &Sec) { +void WinCOFFWriter::writeSection(MCAssembler &Asm, const COFFSection &Sec) { if (Sec.Number == -1) return; @@ -629,7 +624,7 @@ void WinCOFFWriter::writeSection(MCAssembler &Asm, const MCAsmLayout &Layout, assert(W.OS.tell() == Sec.Header.PointerToRawData && "Section::PointerToRawData is insane!"); - uint32_t CRC = writeSectionContents(Asm, Layout, *Sec.MCSection); + uint32_t CRC = writeSectionContents(Asm, *Sec.MCSection); // Update the section definition auxiliary symbol to record the CRC. COFFSymbol::AuxiliarySymbols &AuxSyms = Sec.Symbol->Aux; @@ -761,8 +756,7 @@ void WinCOFFWriter::assignSectionNumbers() { } // Assign file offsets to COFF object file structures. -void WinCOFFWriter::assignFileOffsets(MCAssembler &Asm, - const MCAsmLayout &Layout) { +void WinCOFFWriter::assignFileOffsets(MCAssembler &Asm) { unsigned Offset = W.OS.tell(); Offset += UseBigObj ? COFF::Header32Size : COFF::Header16Size; @@ -1127,7 +1121,7 @@ uint64_t WinCOFFWriter::writeObject(MCAssembler &Asm) { } } - assignFileOffsets(Asm, *Asm.getLayout()); + assignFileOffsets(Asm); // MS LINK expects to be able to use this timestamp to implement their // /INCREMENTAL feature. @@ -1157,7 +1151,7 @@ uint64_t WinCOFFWriter::writeObject(MCAssembler &Asm) { // Write section contents. for (std::unique_ptr<COFFSection> &Sec : Sections) - writeSection(Asm, *Asm.getLayout(), *Sec); + writeSection(Asm, *Sec); assert(W.OS.tell() == Header.PointerToSymbolTable && "Header::PointerToSymbolTable is insane!"); diff --git a/llvm/lib/MC/XCOFFObjectWriter.cpp b/llvm/lib/MC/XCOFFObjectWriter.cpp index bce9ebd..03a4ee8 100644 --- a/llvm/lib/MC/XCOFFObjectWriter.cpp +++ b/llvm/lib/MC/XCOFFObjectWriter.cpp @@ -12,7 +12,6 @@ #include "llvm/BinaryFormat/XCOFF.h" #include "llvm/MC/MCAsmBackend.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCFixup.h" #include "llvm/MC/MCFixupKindInfo.h" @@ -379,18 +378,16 @@ class XCOFFObjectWriter : public MCObjectWriter { void writeSectionHeaderTable(); void writeSections(const MCAssembler &Asm); void writeSectionForControlSectionEntry(const MCAssembler &Asm, - const MCAsmLayout &Layout, const CsectSectionEntry &CsectEntry, uint64_t &CurrentAddressLocation); void writeSectionForDwarfSectionEntry(const MCAssembler &Asm, - const MCAsmLayout &Layout, const DwarfSectionEntry &DwarfEntry, uint64_t &CurrentAddressLocation); - void writeSectionForExceptionSectionEntry( - const MCAssembler &Asm, const MCAsmLayout &Layout, - ExceptionSectionEntry &ExceptionEntry, uint64_t &CurrentAddressLocation); + void + writeSectionForExceptionSectionEntry(const MCAssembler &Asm, + ExceptionSectionEntry &ExceptionEntry, + uint64_t &CurrentAddressLocation); void writeSectionForCInfoSymSectionEntry(const MCAssembler &Asm, - const MCAsmLayout &Layout, CInfoSymSectionEntry &CInfoSymEntry, uint64_t &CurrentAddressLocation); void writeSymbolTable(MCAssembler &Asm); @@ -419,7 +416,7 @@ class XCOFFObjectWriter : public MCObjectWriter { // *) Assigns symbol table indices. // *) Builds up the section header table by adding any non-empty sections to // `Sections`. - void assignAddressesAndIndices(MCAssembler &Asm, const MCAsmLayout &); + void assignAddressesAndIndices(MCAssembler &Asm); // Called after relocations are recorded. void finalizeSectionInfo(); void finalizeRelocationInfo(SectionEntry *Sec, uint64_t RelCount); @@ -655,7 +652,7 @@ void XCOFFObjectWriter::executePostLayoutBinding(MCAssembler &Asm) { Strings.add(Vers); Strings.finalize(); - assignAddressesAndIndices(Asm, *Asm.getLayout()); + assignAddressesAndIndices(Asm); } void XCOFFObjectWriter::recordRelocation(MCAssembler &Asm, @@ -813,17 +810,14 @@ void XCOFFObjectWriter::recordRelocation(MCAssembler &Asm, } void XCOFFObjectWriter::writeSections(const MCAssembler &Asm) { - auto &Layout = *Asm.getLayout(); uint64_t CurrentAddressLocation = 0; for (const auto *Section : Sections) - writeSectionForControlSectionEntry(Asm, Layout, *Section, - CurrentAddressLocation); + writeSectionForControlSectionEntry(Asm, *Section, CurrentAddressLocation); for (const auto &DwarfSection : DwarfSections) - writeSectionForDwarfSectionEntry(Asm, Layout, DwarfSection, - CurrentAddressLocation); - writeSectionForExceptionSectionEntry(Asm, Layout, ExceptionSection, + writeSectionForDwarfSectionEntry(Asm, DwarfSection, CurrentAddressLocation); + writeSectionForExceptionSectionEntry(Asm, ExceptionSection, CurrentAddressLocation); - writeSectionForCInfoSymSectionEntry(Asm, Layout, CInfoSymSection, + writeSectionForCInfoSymSectionEntry(Asm, CInfoSymSection, CurrentAddressLocation); } @@ -1419,8 +1413,7 @@ void XCOFFObjectWriter::addCInfoSymEntry(StringRef Name, StringRef Metadata) { std::make_unique<CInfoSymInfo>(Name.str(), Metadata.str())); } -void XCOFFObjectWriter::assignAddressesAndIndices(MCAssembler &Asm, - const MCAsmLayout &Layout) { +void XCOFFObjectWriter::assignAddressesAndIndices(MCAssembler &Asm) { // The symbol table starts with all the C_FILE symbols. Each C_FILE symbol // requires 1 or 2 auxiliary entries. uint32_t SymbolTableIndex = @@ -1597,8 +1590,8 @@ void XCOFFObjectWriter::assignAddressesAndIndices(MCAssembler &Asm, } void XCOFFObjectWriter::writeSectionForControlSectionEntry( - const MCAssembler &Asm, const MCAsmLayout &Layout, - const CsectSectionEntry &CsectEntry, uint64_t &CurrentAddressLocation) { + const MCAssembler &Asm, const CsectSectionEntry &CsectEntry, + uint64_t &CurrentAddressLocation) { // Nothing to write for this Section. if (CsectEntry.Index == SectionEntry::UninitializedIndex) return; @@ -1644,8 +1637,8 @@ void XCOFFObjectWriter::writeSectionForControlSectionEntry( } void XCOFFObjectWriter::writeSectionForDwarfSectionEntry( - const MCAssembler &Asm, const MCAsmLayout &Layout, - const DwarfSectionEntry &DwarfEntry, uint64_t &CurrentAddressLocation) { + const MCAssembler &Asm, const DwarfSectionEntry &DwarfEntry, + uint64_t &CurrentAddressLocation) { // There could be a gap (without corresponding zero padding) between // sections. For example DWARF section alignment is bigger than // DefaultSectionAlign. @@ -1672,8 +1665,8 @@ void XCOFFObjectWriter::writeSectionForDwarfSectionEntry( } void XCOFFObjectWriter::writeSectionForExceptionSectionEntry( - const MCAssembler &Asm, const MCAsmLayout &Layout, - ExceptionSectionEntry &ExceptionEntry, uint64_t &CurrentAddressLocation) { + const MCAssembler &Asm, ExceptionSectionEntry &ExceptionEntry, + uint64_t &CurrentAddressLocation) { for (auto it = ExceptionEntry.ExceptionTable.begin(); it != ExceptionEntry.ExceptionTable.end(); it++) { // For every symbol that has exception entries, you must start the entries @@ -1695,8 +1688,8 @@ void XCOFFObjectWriter::writeSectionForExceptionSectionEntry( } void XCOFFObjectWriter::writeSectionForCInfoSymSectionEntry( - const MCAssembler &Asm, const MCAsmLayout &Layout, - CInfoSymSectionEntry &CInfoSymEntry, uint64_t &CurrentAddressLocation) { + const MCAssembler &Asm, CInfoSymSectionEntry &CInfoSymEntry, + uint64_t &CurrentAddressLocation) { if (!CInfoSymSection.Entry) return; diff --git a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp index 130d1e96..e0c3cc5 100644 --- a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp +++ b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp @@ -9972,6 +9972,26 @@ SDValue AArch64TargetLowering::LowerCTPOP_PARITY(SDValue Op, Val = DAG.getBitcast(VT8Bit, Val); Val = DAG.getNode(ISD::CTPOP, DL, VT8Bit, Val); + if (Subtarget->hasDotProd() && VT.getScalarSizeInBits() != 16 && + VT.getVectorNumElements() >= 2) { + EVT DT = VT == MVT::v2i64 ? MVT::v4i32 : VT; + SDValue Zeros = DAG.getConstant(0, DL, DT); + SDValue Ones = DAG.getConstant(1, DL, VT8Bit); + + if (VT == MVT::v2i64) { + Val = DAG.getNode(AArch64ISD::UDOT, DL, DT, Zeros, Ones, Val); + Val = DAG.getNode(AArch64ISD::UADDLP, DL, VT, Val); + } else if (VT == MVT::v2i32) { + Val = DAG.getNode(AArch64ISD::UDOT, DL, DT, Zeros, Ones, Val); + } else if (VT == MVT::v4i32) { + Val = DAG.getNode(AArch64ISD::UDOT, DL, DT, Zeros, Ones, Val); + } else { + llvm_unreachable("Unexpected type for custom ctpop lowering"); + } + + return Val; + } + // Widen v8i8/v16i8 CTPOP result to VT by repeatedly widening pairwise adds. unsigned EltSize = 8; unsigned NumElts = VT.is64BitVector() ? 8 : 16; diff --git a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp index c0abbd3..eb60b96 100644 --- a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp +++ b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp @@ -985,6 +985,16 @@ static bool isAllActivePredicate(Value *Pred) { m_ConstantInt<AArch64SVEPredPattern::all>())); } +// Erase unary operation where predicate has all inactive lanes +static std::optional<Instruction *> +instCombineSVENoActiveUnaryErase(InstCombiner &IC, IntrinsicInst &II, + int PredPos) { + if (match(II.getOperand(PredPos), m_ZeroInt())) { + return IC.eraseInstFromFunction(II); + } + return std::nullopt; +} + // Simplify unary operation where predicate has all inactive lanes by replacing // instruction with zeroed object static std::optional<Instruction *> @@ -2007,6 +2017,32 @@ AArch64TTIImpl::instCombineIntrinsic(InstCombiner &IC, default: break; + case Intrinsic::aarch64_sve_st1_scatter: + case Intrinsic::aarch64_sve_st1_scatter_scalar_offset: + case Intrinsic::aarch64_sve_st1_scatter_sxtw: + case Intrinsic::aarch64_sve_st1_scatter_sxtw_index: + case Intrinsic::aarch64_sve_st1_scatter_uxtw: + case Intrinsic::aarch64_sve_st1_scatter_uxtw_index: + case Intrinsic::aarch64_sve_st1dq: + case Intrinsic::aarch64_sve_st1q_scatter_index: + case Intrinsic::aarch64_sve_st1q_scatter_scalar_offset: + case Intrinsic::aarch64_sve_st1q_scatter_vector_offset: + case Intrinsic::aarch64_sve_st1wq: + case Intrinsic::aarch64_sve_stnt1: + case Intrinsic::aarch64_sve_stnt1_scatter: + case Intrinsic::aarch64_sve_stnt1_scatter_index: + case Intrinsic::aarch64_sve_stnt1_scatter_scalar_offset: + case Intrinsic::aarch64_sve_stnt1_scatter_uxtw: + return instCombineSVENoActiveUnaryErase(IC, II, 1); + case Intrinsic::aarch64_sve_st2: + case Intrinsic::aarch64_sve_st2q: + return instCombineSVENoActiveUnaryErase(IC, II, 2); + case Intrinsic::aarch64_sve_st3: + case Intrinsic::aarch64_sve_st3q: + return instCombineSVENoActiveUnaryErase(IC, II, 3); + case Intrinsic::aarch64_sve_st4: + case Intrinsic::aarch64_sve_st4q: + return instCombineSVENoActiveUnaryErase(IC, II, 4); case Intrinsic::aarch64_sve_ld1_gather: case Intrinsic::aarch64_sve_ld1_gather_scalar_offset: case Intrinsic::aarch64_sve_ld1_gather_sxtw: diff --git a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.cpp b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.cpp index 0c5a9d7..fb8eb9f 100644 --- a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.cpp +++ b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.cpp @@ -101,9 +101,9 @@ MCFragment *AArch64MCExpr::findAssociatedFragment() const { } bool AArch64MCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { - if (!getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup)) return false; Res = @@ -187,9 +187,9 @@ MCFragment *AArch64AuthMCExpr::findAssociatedFragment() const { } bool AArch64AuthMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { - if (!getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup)) return false; if (Res.getSymB()) diff --git a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.h b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.h index 4823598..cf3a90f 100644 --- a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.h +++ b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCExpr.h @@ -167,7 +167,7 @@ public: MCFragment *findAssociatedFragment() const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void fixELFSymbolsInTLSFixups(MCAssembler &Asm) const override; @@ -201,7 +201,7 @@ public: MCFragment *findAssociatedFragment() const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; static bool classof(const MCExpr *E) { diff --git a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MachObjectWriter.cpp b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MachObjectWriter.cpp index d5177aa..33f0ca2 100644 --- a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MachObjectWriter.cpp +++ b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MachObjectWriter.cpp @@ -12,7 +12,6 @@ #include "llvm/BinaryFormat/MachO.h" #include "llvm/MC/MCAsmInfo.h" #include "llvm/MC/MCAsmInfoDarwin.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCExpr.h" @@ -278,18 +277,14 @@ void AArch64MachObjectWriter::recordRelocation( return; } - Value += - (!A->getFragment() ? 0 - : Writer->getSymbolAddress(*A, *Asm.getLayout())) - - (!A_Base || !A_Base->getFragment() - ? 0 - : Writer->getSymbolAddress(*A_Base, *Asm.getLayout())); - Value -= - (!B->getFragment() ? 0 - : Writer->getSymbolAddress(*B, *Asm.getLayout())) - - (!B_Base || !B_Base->getFragment() - ? 0 - : Writer->getSymbolAddress(*B_Base, *Asm.getLayout())); + Value += (!A->getFragment() ? 0 : Writer->getSymbolAddress(*A, Asm)) - + (!A_Base || !A_Base->getFragment() + ? 0 + : Writer->getSymbolAddress(*A_Base, Asm)); + Value -= (!B->getFragment() ? 0 : Writer->getSymbolAddress(*B, Asm)) - + (!B_Base || !B_Base->getFragment() + ? 0 + : Writer->getSymbolAddress(*B_Base, Asm)); Type = MachO::ARM64_RELOC_UNSIGNED; @@ -358,7 +353,7 @@ void AArch64MachObjectWriter::recordRelocation( // The index is the section ordinal (1-based). const MCSection &Sec = Symbol->getSection(); Index = Sec.getOrdinal() + 1; - Value += Writer->getSymbolAddress(*Symbol, *Asm.getLayout()); + Value += Writer->getSymbolAddress(*Symbol, Asm); if (IsPCRel) Value -= Writer->getFragmentAddress(Asm, Fragment) + Fixup.getOffset() + diff --git a/llvm/lib/Target/AMDGPU/AMDGPU.td b/llvm/lib/Target/AMDGPU/AMDGPU.td index 63d8334..2f3890e 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPU.td +++ b/llvm/lib/Target/AMDGPU/AMDGPU.td @@ -2013,6 +2013,7 @@ def HasVOP3PInsts : Predicate<"Subtarget->hasVOP3PInsts()">, AssemblerPredicate<(all_of FeatureVOP3P)>; def NotHasMed3_16 : Predicate<"!Subtarget->hasMed3_16()">; +def HasMed3_16 : Predicate<"Subtarget->hasMed3_16()">; def HasMinMaxDenormModes : Predicate<"Subtarget->supportsMinMaxDenormModes()">; def NotHasMinMaxDenormModes : Predicate<"!Subtarget->supportsMinMaxDenormModes()">; diff --git a/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp b/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp index 8062bc1..08c1cd5 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp +++ b/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp @@ -651,9 +651,12 @@ static Constant *getIdentityValueForAtomicOp(Type *const Ty, case AtomicRMWInst::FSub: return ConstantFP::get(C, APFloat::getZero(Ty->getFltSemantics(), false)); case AtomicRMWInst::FMin: - return ConstantFP::get(C, APFloat::getInf(Ty->getFltSemantics(), false)); case AtomicRMWInst::FMax: - return ConstantFP::get(C, APFloat::getInf(Ty->getFltSemantics(), true)); + // FIXME: atomicrmw fmax/fmin behave like llvm.maxnum/minnum so NaN is the + // closest thing they have to an identity, but it still does not preserve + // the difference between quiet and signaling NaNs or NaNs with different + // payloads. + return ConstantFP::get(C, APFloat::getNaN(Ty->getFltSemantics())); } } diff --git a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.cpp b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.cpp index 83fbf4a..b2a9667 100644 --- a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.cpp +++ b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.cpp @@ -10,6 +10,7 @@ #include "GCNSubtarget.h" #include "Utils/AMDGPUBaseInfo.h" #include "llvm/IR/Function.h" +#include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCStreamer.h" #include "llvm/MC/MCSymbol.h" @@ -93,12 +94,11 @@ static int64_t op(AMDGPUMCExpr::VariantKind Kind, int64_t Arg1, int64_t Arg2) { } } -bool AMDGPUMCExpr::evaluateExtraSGPRs(MCValue &Res, const MCAsmLayout *Layout, +bool AMDGPUMCExpr::evaluateExtraSGPRs(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const { auto TryGetMCExprValue = [&](const MCExpr *Arg, uint64_t &ConstantValue) { MCValue MCVal; - if (!Arg->evaluateAsRelocatable(MCVal, Layout, Fixup) || - !MCVal.isAbsolute()) + if (!Arg->evaluateAsRelocatable(MCVal, Asm, Fixup) || !MCVal.isAbsolute()) return false; ConstantValue = MCVal.getConstant(); @@ -123,12 +123,11 @@ bool AMDGPUMCExpr::evaluateExtraSGPRs(MCValue &Res, const MCAsmLayout *Layout, return true; } -bool AMDGPUMCExpr::evaluateTotalNumVGPR(MCValue &Res, const MCAsmLayout *Layout, +bool AMDGPUMCExpr::evaluateTotalNumVGPR(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const { auto TryGetMCExprValue = [&](const MCExpr *Arg, uint64_t &ConstantValue) { MCValue MCVal; - if (!Arg->evaluateAsRelocatable(MCVal, Layout, Fixup) || - !MCVal.isAbsolute()) + if (!Arg->evaluateAsRelocatable(MCVal, Asm, Fixup) || !MCVal.isAbsolute()) return false; ConstantValue = MCVal.getConstant(); @@ -151,12 +150,11 @@ bool AMDGPUMCExpr::evaluateTotalNumVGPR(MCValue &Res, const MCAsmLayout *Layout, return true; } -bool AMDGPUMCExpr::evaluateAlignTo(MCValue &Res, const MCAsmLayout *Layout, +bool AMDGPUMCExpr::evaluateAlignTo(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const { auto TryGetMCExprValue = [&](const MCExpr *Arg, uint64_t &ConstantValue) { MCValue MCVal; - if (!Arg->evaluateAsRelocatable(MCVal, Layout, Fixup) || - !MCVal.isAbsolute()) + if (!Arg->evaluateAsRelocatable(MCVal, Asm, Fixup) || !MCVal.isAbsolute()) return false; ConstantValue = MCVal.getConstant(); @@ -173,12 +171,11 @@ bool AMDGPUMCExpr::evaluateAlignTo(MCValue &Res, const MCAsmLayout *Layout, return true; } -bool AMDGPUMCExpr::evaluateOccupancy(MCValue &Res, const MCAsmLayout *Layout, +bool AMDGPUMCExpr::evaluateOccupancy(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const { auto TryGetMCExprValue = [&](const MCExpr *Arg, uint64_t &ConstantValue) { MCValue MCVal; - if (!Arg->evaluateAsRelocatable(MCVal, Layout, Fixup) || - !MCVal.isAbsolute()) + if (!Arg->evaluateAsRelocatable(MCVal, Asm, Fixup) || !MCVal.isAbsolute()) return false; ConstantValue = MCVal.getConstant(); @@ -218,27 +215,25 @@ bool AMDGPUMCExpr::evaluateOccupancy(MCValue &Res, const MCAsmLayout *Layout, } bool AMDGPUMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { std::optional<int64_t> Total; - switch (Kind) { default: break; case AGVK_ExtraSGPRs: - return evaluateExtraSGPRs(Res, Layout, Fixup); + return evaluateExtraSGPRs(Res, Asm, Fixup); case AGVK_AlignTo: - return evaluateAlignTo(Res, Layout, Fixup); + return evaluateAlignTo(Res, Asm, Fixup); case AGVK_TotalNumVGPRs: - return evaluateTotalNumVGPR(Res, Layout, Fixup); + return evaluateTotalNumVGPR(Res, Asm, Fixup); case AGVK_Occupancy: - return evaluateOccupancy(Res, Layout, Fixup); + return evaluateOccupancy(Res, Asm, Fixup); } for (const MCExpr *Arg : Args) { MCValue ArgRes; - if (!Arg->evaluateAsRelocatable(ArgRes, Layout, Fixup) || - !ArgRes.isAbsolute()) + if (!Arg->evaluateAsRelocatable(ArgRes, Asm, Fixup) || !ArgRes.isAbsolute()) return false; if (!Total.has_value()) diff --git a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.h b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.h index 207a619..970802d 100644 --- a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.h +++ b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCExpr.h @@ -48,13 +48,13 @@ private: AMDGPUMCExpr(VariantKind Kind, ArrayRef<const MCExpr *> Args, MCContext &Ctx); ~AMDGPUMCExpr(); - bool evaluateExtraSGPRs(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateExtraSGPRs(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const; - bool evaluateTotalNumVGPR(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateTotalNumVGPR(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const; - bool evaluateAlignTo(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAlignTo(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const; - bool evaluateOccupancy(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateOccupancy(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const; public: @@ -94,7 +94,7 @@ public: const MCExpr *getSubExpr(size_t Index) const; void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override; diff --git a/llvm/lib/Target/AMDGPU/SIInstructions.td b/llvm/lib/Target/AMDGPU/SIInstructions.td index 835f44f..77b17a0 100644 --- a/llvm/lib/Target/AMDGPU/SIInstructions.td +++ b/llvm/lib/Target/AMDGPU/SIInstructions.td @@ -3534,16 +3534,6 @@ multiclass FPMed3Pat<ValueType vt, DSTCLAMP.NONE, DSTOMOD.NONE)>; } -class FP16Med3Pat<ValueType vt, - Instruction med3Inst> : GCNPat< - (fmaxnum_like_nnan (fminnum_like (VOP3Mods vt:$src0, i32:$src0_mods), - (VOP3Mods vt:$src1, i32:$src1_mods)), - (fminnum_like (fmaxnum_like (VOP3Mods vt:$src0, i32:$src0_mods), - (VOP3Mods vt:$src1, i32:$src1_mods)), - (vt (VOP3Mods vt:$src2, i32:$src2_mods)))), - (med3Inst $src0_mods, $src0, $src1_mods, $src1, $src2_mods, $src2, DSTCLAMP.NONE) ->; - multiclass Int16Med3Pat<Instruction med3Inst, SDPatternOperator min, SDPatternOperator max> { @@ -3566,6 +3556,10 @@ multiclass Int16Med3Pat<Instruction med3Inst, defm : FPMed3Pat<f32, V_MED3_F32_e64>; +let SubtargetPredicate = HasMed3_16 in { +defm : FPMed3Pat<f16, V_MED3_F16_e64>; +} + class IntMinMaxPat<Instruction minmaxInst, SDPatternOperator min_or_max, SDPatternOperator max_or_min_oneuse> : AMDGPUPat < @@ -3611,7 +3605,6 @@ def : FPMinCanonMaxPat<V_MAXMIN_F16_e64, f16, fminnum_like, fmaxnum_like_oneuse> } let OtherPredicates = [isGFX9Plus] in { -def : FP16Med3Pat<f16, V_MED3_F16_e64>; defm : Int16Med3Pat<V_MED3_I16_e64, smin, smax>; defm : Int16Med3Pat<V_MED3_U16_e64, umin, umax>; } // End Predicates = [isGFX9Plus] diff --git a/llvm/lib/Target/ARM/MCTargetDesc/ARMAsmBackend.cpp b/llvm/lib/Target/ARM/MCTargetDesc/ARMAsmBackend.cpp index 48b4b69..33eb1cf 100644 --- a/llvm/lib/Target/ARM/MCTargetDesc/ARMAsmBackend.cpp +++ b/llvm/lib/Target/ARM/MCTargetDesc/ARMAsmBackend.cpp @@ -17,7 +17,6 @@ #include "llvm/BinaryFormat/ELF.h" #include "llvm/BinaryFormat/MachO.h" #include "llvm/MC/MCAsmBackend.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCDirectives.h" diff --git a/llvm/lib/Target/ARM/MCTargetDesc/ARMMCExpr.h b/llvm/lib/Target/ARM/MCTargetDesc/ARMMCExpr.h index edeff9c..5f32b37 100644 --- a/llvm/lib/Target/ARM/MCTargetDesc/ARMMCExpr.h +++ b/llvm/lib/Target/ARM/MCTargetDesc/ARMMCExpr.h @@ -81,8 +81,7 @@ public: /// @} void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override { return false; } diff --git a/llvm/lib/Target/ARM/MCTargetDesc/ARMMachObjectWriter.cpp b/llvm/lib/Target/ARM/MCTargetDesc/ARMMachObjectWriter.cpp index 2a2372d..e4a2cce 100644 --- a/llvm/lib/Target/ARM/MCTargetDesc/ARMMachObjectWriter.cpp +++ b/llvm/lib/Target/ARM/MCTargetDesc/ARMMachObjectWriter.cpp @@ -12,7 +12,6 @@ #include "llvm/ADT/StringExtras.h" #include "llvm/ADT/Twine.h" #include "llvm/BinaryFormat/MachO.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCExpr.h" @@ -158,7 +157,7 @@ void ARMMachObjectWriter::recordARMScatteredHalfRelocation( return; } - uint32_t Value = Writer->getSymbolAddress(*A, *Asm.getLayout()); + uint32_t Value = Writer->getSymbolAddress(*A, Asm); uint32_t Value2 = 0; uint64_t SecAddr = Writer->getSectionAddress(A->getFragment()->getParent()); FixedValue += SecAddr; @@ -175,7 +174,7 @@ void ARMMachObjectWriter::recordARMScatteredHalfRelocation( // Select the appropriate difference relocation type. Type = MachO::ARM_RELOC_HALF_SECTDIFF; - Value2 = Writer->getSymbolAddress(B->getSymbol(), *Asm.getLayout()); + Value2 = Writer->getSymbolAddress(B->getSymbol(), Asm); FixedValue -= Writer->getSectionAddress(SB->getFragment()->getParent()); } @@ -267,7 +266,7 @@ void ARMMachObjectWriter::recordARMScatteredRelocation( return; } - uint32_t Value = Writer->getSymbolAddress(*A, *Asm.getLayout()); + uint32_t Value = Writer->getSymbolAddress(*A, Asm); uint64_t SecAddr = Writer->getSectionAddress(A->getFragment()->getParent()); FixedValue += SecAddr; uint32_t Value2 = 0; @@ -285,7 +284,7 @@ void ARMMachObjectWriter::recordARMScatteredRelocation( // Select the appropriate difference relocation type. Type = MachO::ARM_RELOC_SECTDIFF; - Value2 = Writer->getSymbolAddress(B->getSymbol(), *Asm.getLayout()); + Value2 = Writer->getSymbolAddress(B->getSymbol(), Asm); FixedValue -= Writer->getSectionAddress(SB->getFragment()->getParent()); } diff --git a/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.cpp b/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.cpp index d6b6546..5386df7 100644 --- a/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.cpp +++ b/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.cpp @@ -8,7 +8,6 @@ #include "AVRMCExpr.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCStreamer.h" @@ -69,10 +68,10 @@ bool AVRMCExpr::evaluateAsConstant(int64_t &Result) const { } bool AVRMCExpr::evaluateAsRelocatableImpl(MCValue &Result, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { MCValue Value; - bool isRelocatable = SubExpr->evaluateAsRelocatable(Value, Layout, Fixup); + bool isRelocatable = SubExpr->evaluateAsRelocatable(Value, Asm, Fixup); if (!isRelocatable) return false; @@ -80,10 +79,10 @@ bool AVRMCExpr::evaluateAsRelocatableImpl(MCValue &Result, if (Value.isAbsolute()) { Result = MCValue::get(evaluateAsInt64(Value.getConstant())); } else { - if (!Layout) + if (!Asm || !Asm->hasLayout()) return false; - MCContext &Context = Layout->getAssembler().getContext(); + MCContext &Context = Asm->getContext(); const MCSymbolRefExpr *Sym = Value.getSymA(); MCSymbolRefExpr::VariantKind Modifier = Sym->getKind(); if (Modifier != MCSymbolRefExpr::VK_None) diff --git a/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.h b/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.h index 5bf6c1a..03d3363 100644 --- a/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.h +++ b/llvm/lib/Target/AVR/MCTargetDesc/AVRMCExpr.h @@ -56,7 +56,7 @@ public: void setNegated(bool negated = true) { Negated = negated; } void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &streamer) const override; diff --git a/llvm/lib/Target/CSKY/MCTargetDesc/CSKYAsmBackend.cpp b/llvm/lib/Target/CSKY/MCTargetDesc/CSKYAsmBackend.cpp index 222dc2c..dd06971 100644 --- a/llvm/lib/Target/CSKY/MCTargetDesc/CSKYAsmBackend.cpp +++ b/llvm/lib/Target/CSKY/MCTargetDesc/CSKYAsmBackend.cpp @@ -9,7 +9,6 @@ #include "CSKYAsmBackend.h" #include "MCTargetDesc/CSKYMCTargetDesc.h" #include "llvm/ADT/DenseMap.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCFixupKindInfo.h" diff --git a/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.cpp b/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.cpp index b998982..d96b700 100644 --- a/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.cpp +++ b/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.cpp @@ -109,10 +109,9 @@ void CSKYMCExpr::fixELFSymbolsInTLSFixups(MCAssembler &Asm) const { fixELFSymbolsInTLSFixupsImpl(getSubExpr(), Asm); } -bool CSKYMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, +bool CSKYMCExpr::evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const { - if (!getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup)) return false; // Some custom fixup types are not valid with symbol difference expressions diff --git a/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.h b/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.h index 9e5b4ca..73aebc7 100644 --- a/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.h +++ b/llvm/lib/Target/CSKY/MCTargetDesc/CSKYMCExpr.h @@ -55,7 +55,7 @@ public: void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; diff --git a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonAsmBackend.cpp b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonAsmBackend.cpp index 2ce91bd..6acc37e 100644 --- a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonAsmBackend.cpp +++ b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonAsmBackend.cpp @@ -14,7 +14,6 @@ #include "MCTargetDesc/HexagonMCShuffler.h" #include "MCTargetDesc/HexagonMCTargetDesc.h" #include "llvm/MC/MCAsmBackend.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCELFObjectWriter.h" @@ -703,8 +702,7 @@ public: return true; } - void finishLayout(MCAssembler const &Asm, - MCAsmLayout &Layout) const override { + void finishLayout(MCAssembler const &Asm) const override { SmallVector<MCFragment *> Frags; for (MCSection &Sec : Asm) { Frags.clear(); diff --git a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.cpp b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.cpp index 0cfea77..598d1e8 100644 --- a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.cpp +++ b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.cpp @@ -25,9 +25,9 @@ HexagonMCExpr *HexagonMCExpr::create(MCExpr const *Expr, MCContext &Ctx) { } bool HexagonMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - MCAsmLayout const *Layout, + const MCAssembler *Asm, MCFixup const *Fixup) const { - return Expr->evaluateAsRelocatable(Res, Layout, Fixup); + return Expr->evaluateAsRelocatable(Res, Asm, Fixup); } void HexagonMCExpr::visitUsedExpr(MCStreamer &Streamer) const { diff --git a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.h b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.h index 6438ac9..eda96e2 100644 --- a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.h +++ b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCExpr.h @@ -16,7 +16,7 @@ class HexagonMCExpr : public MCTargetExpr { public: static HexagonMCExpr *create(MCExpr const *Expr, MCContext &Ctx); void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override; diff --git a/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.cpp b/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.cpp index 56d5fbf..285e938 100644 --- a/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.cpp +++ b/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.cpp @@ -47,9 +47,9 @@ void LanaiMCExpr::visitUsedExpr(MCStreamer &Streamer) const { } bool LanaiMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { - if (!getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup)) return false; Res = diff --git a/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.h b/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.h index c99af32..1829165 100644 --- a/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.h +++ b/llvm/lib/Target/Lanai/MCTargetDesc/LanaiMCExpr.h @@ -36,7 +36,7 @@ public: const MCExpr *getSubExpr() const { return Expr; } void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override { diff --git a/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchAsmBackend.cpp b/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchAsmBackend.cpp index e3200b0..0c24008 100644 --- a/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchAsmBackend.cpp +++ b/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchAsmBackend.cpp @@ -13,7 +13,6 @@ #include "LoongArchAsmBackend.h" #include "LoongArchFixupKinds.h" #include "llvm/MC/MCAsmInfo.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCELFObjectWriter.h" @@ -303,7 +302,7 @@ std::pair<bool, bool> LoongArchAsmBackend::relaxLEB128(const MCAssembler &Asm, MCLEBFragment &LF, int64_t &Value) const { const MCExpr &Expr = LF.getValue(); - if (LF.isSigned() || !Expr.evaluateKnownAbsolute(Value, *Asm.getLayout())) + if (LF.isSigned() || !Expr.evaluateKnownAbsolute(Value, Asm)) return std::make_pair(false, false); LF.getFixups().push_back( MCFixup::create(0, &Expr, FK_Data_leb128, Expr.getLoc())); @@ -313,7 +312,6 @@ std::pair<bool, bool> LoongArchAsmBackend::relaxLEB128(const MCAssembler &Asm, bool LoongArchAsmBackend::relaxDwarfLineAddr(const MCAssembler &Asm, MCDwarfLineAddrFragment &DF, bool &WasRelaxed) const { - auto &Layout = *Asm.getLayout(); MCContext &C = Asm.getContext(); int64_t LineDelta = DF.getLineDelta(); @@ -323,9 +321,9 @@ bool LoongArchAsmBackend::relaxDwarfLineAddr(const MCAssembler &Asm, size_t OldSize = Data.size(); int64_t Value; - if (AddrDelta.evaluateAsAbsolute(Value, Layout)) + if (AddrDelta.evaluateAsAbsolute(Value, Asm)) return false; - bool IsAbsolute = AddrDelta.evaluateKnownAbsolute(Value, Layout); + bool IsAbsolute = AddrDelta.evaluateKnownAbsolute(Value, Asm); assert(IsAbsolute && "CFA with invalid expression"); (void)IsAbsolute; @@ -387,11 +385,10 @@ bool LoongArchAsmBackend::relaxDwarfCFA(const MCAssembler &Asm, SmallVectorImpl<MCFixup> &Fixups = DF.getFixups(); size_t OldSize = Data.size(); - auto &Layout = *Asm.getLayout(); int64_t Value; - if (AddrDelta.evaluateAsAbsolute(Value, Layout)) + if (AddrDelta.evaluateAsAbsolute(Value, Asm)) return false; - bool IsAbsolute = AddrDelta.evaluateKnownAbsolute(Value, Layout); + bool IsAbsolute = AddrDelta.evaluateKnownAbsolute(Value, Asm); assert(IsAbsolute && "CFA with invalid expression"); (void)IsAbsolute; @@ -399,10 +396,8 @@ bool LoongArchAsmBackend::relaxDwarfCFA(const MCAssembler &Asm, Fixups.clear(); raw_svector_ostream OS(Data); - assert( - Layout.getAssembler().getContext().getAsmInfo()->getMinInstAlignment() == - 1 && - "expected 1-byte alignment"); + assert(Asm.getContext().getAsmInfo()->getMinInstAlignment() == 1 && + "expected 1-byte alignment"); if (Value == 0) { WasRelaxed = OldSize != Data.size(); return true; diff --git a/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.cpp b/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.cpp index 34f9bc6..9b53779 100644 --- a/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.cpp +++ b/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.cpp @@ -44,7 +44,7 @@ void LoongArchMCExpr::printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const { } bool LoongArchMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { // Explicitly drop the layout and assembler to prevent any symbolic folding in // the expression handling. This is required to preserve symbolic difference diff --git a/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.h b/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.h index 71dd5bd..148fea3 100644 --- a/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.h +++ b/llvm/lib/Target/LoongArch/MCTargetDesc/LoongArchMCExpr.h @@ -92,7 +92,7 @@ public: bool getRelaxHint() const { return RelaxHint; } void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override { diff --git a/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.cpp b/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.cpp index a25783b..c856aa8 100644 --- a/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.cpp +++ b/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.cpp @@ -129,16 +129,14 @@ void MipsMCExpr::printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const { OS << ')'; } -bool -MipsMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, - const MCFixup *Fixup) const { +bool MipsMCExpr::evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, + const MCFixup *Fixup) const { // Look for the %hi(%neg(%gp_rel(X))) and %lo(%neg(%gp_rel(X))) special cases. if (isGpOff()) { const MCExpr *SubExpr = cast<MipsMCExpr>(cast<MipsMCExpr>(getSubExpr())->getSubExpr()) ->getSubExpr(); - if (!SubExpr->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!SubExpr->evaluateAsRelocatable(Res, Asm, Fixup)) return false; Res = MCValue::get(Res.getSymA(), Res.getSymB(), Res.getConstant(), @@ -146,7 +144,7 @@ MipsMCExpr::evaluateAsRelocatableImpl(MCValue &Res, return true; } - if (!getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup)) return false; if (Res.getRefKind() != MCSymbolRefExpr::VK_None) @@ -164,7 +162,7 @@ MipsMCExpr::evaluateAsRelocatableImpl(MCValue &Res, case MEK_DTPREL: // MEK_DTPREL is used for marking TLS DIEExpr only // and contains a regular sub-expression. - return getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup); + return getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup); case MEK_DTPREL_HI: case MEK_DTPREL_LO: case MEK_GOT: diff --git a/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.h b/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.h index edc12e8..fc97200 100644 --- a/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.h +++ b/llvm/lib/Target/Mips/MCTargetDesc/MipsMCExpr.h @@ -9,7 +9,6 @@ #ifndef LLVM_LIB_TARGET_MIPS_MCTARGETDESC_MIPSMCEXPR_H #define LLVM_LIB_TARGET_MIPS_MCTARGETDESC_MIPSMCEXPR_H -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCExpr.h" #include "llvm/MC/MCValue.h" @@ -67,7 +66,7 @@ public: const MCExpr *getSubExpr() const { return Expr; } void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; diff --git a/llvm/lib/Target/NVPTX/MCTargetDesc/NVPTXInstPrinter.cpp b/llvm/lib/Target/NVPTX/MCTargetDesc/NVPTXInstPrinter.cpp index b7a20c3..380d878 100644 --- a/llvm/lib/Target/NVPTX/MCTargetDesc/NVPTXInstPrinter.cpp +++ b/llvm/lib/Target/NVPTX/MCTargetDesc/NVPTXInstPrinter.cpp @@ -60,6 +60,9 @@ void NVPTXInstPrinter::printRegName(raw_ostream &OS, MCRegister Reg) const { case 6: OS << "%fd"; break; + case 7: + OS << "%rq"; + break; } unsigned VReg = Reg.id() & 0x0FFFFFFF; diff --git a/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp b/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp index ca077d4..1645261 100644 --- a/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp +++ b/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp @@ -315,6 +315,8 @@ unsigned NVPTXAsmPrinter::encodeVirtualRegister(unsigned Reg) { Ret = (5 << 28); } else if (RC == &NVPTX::Float64RegsRegClass) { Ret = (6 << 28); + } else if (RC == &NVPTX::Int128RegsRegClass) { + Ret = (7 << 28); } else { report_fatal_error("Bad register class"); } diff --git a/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.cpp b/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.cpp index 1e1cbb1..11193c1 100644 --- a/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.cpp +++ b/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.cpp @@ -519,6 +519,20 @@ void NVPTXDAGToDAGISel::Select(SDNode *N) { if (tryConstantFP(N)) return; break; + case ISD::CopyToReg: { + if (N->getOperand(1).getValueType() == MVT::i128) { + SelectV2I64toI128(N); + return; + } + break; + } + case ISD::CopyFromReg: { + if (N->getOperand(1).getValueType() == MVT::i128) { + SelectI128toV2I64(N); + return; + } + break; + } default: break; } @@ -3798,6 +3812,60 @@ bool NVPTXDAGToDAGISel::SelectInlineAsmMemoryOperand( return true; } +void NVPTXDAGToDAGISel::SelectV2I64toI128(SDNode *N) { + // Lower a CopyToReg with two 64-bit inputs + // Dst:i128, lo:i64, hi:i64 + // + // CopyToReg Dst, lo, hi; + // + // ==> + // + // tmp = V2I64toI128 {lo, hi}; + // CopyToReg Dst, tmp; + SDValue Dst = N->getOperand(1); + SDValue Lo = N->getOperand(2); + SDValue Hi = N->getOperand(3); + + SDLoc DL(N); + SDNode *Mov = + CurDAG->getMachineNode(NVPTX::V2I64toI128, DL, MVT::i128, {Lo, Hi}); + + SmallVector<SDValue, 4> NewOps(N->getNumOperands() - 1); + NewOps[0] = N->getOperand(0); + NewOps[1] = Dst; + NewOps[2] = SDValue(Mov, 0); + if (N->getNumOperands() == 5) + NewOps[3] = N->getOperand(4); + SDValue NewValue = CurDAG->getNode(ISD::CopyToReg, DL, SmallVector<EVT>(N->values()), NewOps); + + ReplaceNode(N, NewValue.getNode()); +} + +void NVPTXDAGToDAGISel::SelectI128toV2I64(SDNode *N) { + // Lower CopyFromReg from a 128-bit regs to two 64-bit regs + // Dst:i128, Src:i128 + // + // {lo, hi} = CopyFromReg Src + // + // ==> + // + // {lo, hi} = I128toV2I64 Src + // + SDValue Ch = N->getOperand(0); + SDValue Src = N->getOperand(1); + SDValue Glue = N->getOperand(2); + SDLoc DL(N); + + // Add Glue and Ch to the operands and results to avoid break the execution + // order + SDNode *Mov = CurDAG->getMachineNode( + NVPTX::I128toV2I64, DL, + {MVT::i64, MVT::i64, Ch.getValueType(), Glue.getValueType()}, + {Src, Ch, Glue}); + + ReplaceNode(N, Mov); +} + /// GetConvertOpcode - Returns the CVT_ instruction opcode that implements a /// conversion from \p SrcTy to \p DestTy. unsigned NVPTXDAGToDAGISel::GetConvertOpcode(MVT DestTy, MVT SrcTy, diff --git a/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.h b/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.h index c552435..49626d4 100644 --- a/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.h +++ b/llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.h @@ -74,7 +74,8 @@ private: bool SelectSETP_F16X2(SDNode *N); bool SelectSETP_BF16X2(SDNode *N); bool tryEXTRACT_VECTOR_ELEMENT(SDNode *N); - + void SelectV2I64toI128(SDNode *N); + void SelectI128toV2I64(SDNode *N); inline SDValue getI32Imm(unsigned Imm, const SDLoc &DL) { return CurDAG->getTargetConstant(Imm, DL, MVT::i32); } diff --git a/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp b/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp index 476a532..26c16ee 100644 --- a/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp +++ b/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp @@ -859,6 +859,10 @@ NVPTXTargetLowering::NVPTXTargetLowering(const NVPTXTargetMachine &TM, setBF16OperationAction(Op, MVT::v2bf16, Legal, Expand); } + // Custom lowering for inline asm with 128-bit operands + setOperationAction(ISD::CopyToReg, MVT::i128, Custom); + setOperationAction(ISD::CopyFromReg, MVT::i128, Custom); + // No FEXP2, FLOG2. The PTX ex2 and log2 functions are always approximate. // No FPOW or FREM in PTX. @@ -2804,6 +2808,8 @@ NVPTXTargetLowering::LowerOperation(SDValue Op, SelectionDAG &DAG) const { return LowerVectorArith(Op, DAG); case ISD::DYNAMIC_STACKALLOC: return LowerDYNAMIC_STACKALLOC(Op, DAG); + case ISD::CopyToReg: + return LowerCopyToReg_128(Op, DAG); default: llvm_unreachable("Custom lowering not defined for operation"); } @@ -3094,6 +3100,54 @@ SDValue NVPTXTargetLowering::LowerSTOREi1(SDValue Op, SelectionDAG &DAG) const { return Result; } +SDValue NVPTXTargetLowering::LowerCopyToReg_128(SDValue Op, + SelectionDAG &DAG) const { + // Change the CopyToReg to take in two 64-bit operands instead of a 128-bit + // operand so that it can pass the legalization. + + assert(Op.getOperand(1).getValueType() == MVT::i128 && + "Custom lowering for 128-bit CopyToReg only"); + + SDNode *Node = Op.getNode(); + SDLoc DL(Node); + + SDValue Cast = DAG.getBitcast(MVT::v2i64, Op->getOperand(2)); + SDValue Lo = DAG.getNode(ISD::EXTRACT_VECTOR_ELT, DL, MVT::i64, Cast, + DAG.getIntPtrConstant(0, DL)); + SDValue Hi = DAG.getNode(ISD::EXTRACT_VECTOR_ELT, DL, MVT::i64, Cast, + DAG.getIntPtrConstant(1, DL)); + + SmallVector<SDValue, 5> NewOps(Op->getNumOperands() + 1); + SmallVector<EVT, 3> ResultsType(Node->values()); + + NewOps[0] = Op->getOperand(0); // Chain + NewOps[1] = Op->getOperand(1); // Dst Reg + NewOps[2] = Lo; // Lower 64-bit + NewOps[3] = Hi; // Higher 64-bit + if (Op.getNumOperands() == 4) + NewOps[4] = Op->getOperand(3); // Glue if exists + + return DAG.getNode(ISD::CopyToReg, DL, ResultsType, NewOps); +} + +unsigned NVPTXTargetLowering::getNumRegisters( + LLVMContext &Context, EVT VT, + std::optional<MVT> RegisterVT = std::nullopt) const { + if (VT == MVT::i128 && RegisterVT == MVT::i128) + return 1; + return TargetLoweringBase::getNumRegisters(Context, VT, RegisterVT); +} + +bool NVPTXTargetLowering::splitValueIntoRegisterParts( + SelectionDAG &DAG, const SDLoc &DL, SDValue Val, SDValue *Parts, + unsigned NumParts, MVT PartVT, std::optional<CallingConv::ID> CC) const { + if (Val.getValueType() == MVT::i128 && NumParts == 1) { + Parts[0] = Val; + return true; + } + return false; +} + // This creates target external symbol for a function parameter. // Name of the symbol is composed from its index and the function name. // Negative index corresponds to special parameter (unsized array) used for @@ -5150,6 +5204,7 @@ NVPTXTargetLowering::getConstraintType(StringRef Constraint) const { case 'l': case 'f': case 'd': + case 'q': case '0': case 'N': return C_RegisterClass; @@ -5175,6 +5230,12 @@ NVPTXTargetLowering::getRegForInlineAsmConstraint(const TargetRegisterInfo *TRI, case 'l': case 'N': return std::make_pair(0U, &NVPTX::Int64RegsRegClass); + case 'q': { + if (STI.getSmVersion() < 70) + report_fatal_error("Inline asm with 128 bit operands is only " + "supported for sm_70 and higher!"); + return std::make_pair(0U, &NVPTX::Int128RegsRegClass); + } case 'f': return std::make_pair(0U, &NVPTX::Float32RegsRegClass); case 'd': @@ -6261,6 +6322,30 @@ static void ReplaceINTRINSIC_W_CHAIN(SDNode *N, SelectionDAG &DAG, } } +static void ReplaceCopyFromReg_128(SDNode *N, SelectionDAG &DAG, + SmallVectorImpl<SDValue> &Results) { + // Change the CopyFromReg to output 2 64-bit results instead of a 128-bit + // result so that it can pass the legalization + SDLoc DL(N); + SDValue Chain = N->getOperand(0); + SDValue Reg = N->getOperand(1); + SDValue Glue = N->getOperand(2); + + assert(Reg.getValueType() == MVT::i128 && + "Custom lowering for CopyFromReg with 128-bit reg only"); + SmallVector<EVT, 4> ResultsType = {MVT::i64, MVT::i64, N->getValueType(1), + N->getValueType(2)}; + SmallVector<SDValue, 3> NewOps = {Chain, Reg, Glue}; + + SDValue NewValue = DAG.getNode(ISD::CopyFromReg, DL, ResultsType, NewOps); + SDValue Pair = DAG.getNode(ISD::BUILD_PAIR, DL, MVT::i128, + {NewValue.getValue(0), NewValue.getValue(1)}); + + Results.push_back(Pair); + Results.push_back(NewValue.getValue(2)); + Results.push_back(NewValue.getValue(3)); +} + void NVPTXTargetLowering::ReplaceNodeResults( SDNode *N, SmallVectorImpl<SDValue> &Results, SelectionDAG &DAG) const { switch (N->getOpcode()) { @@ -6272,6 +6357,9 @@ void NVPTXTargetLowering::ReplaceNodeResults( case ISD::INTRINSIC_W_CHAIN: ReplaceINTRINSIC_W_CHAIN(N, DAG, Results); return; + case ISD::CopyFromReg: + ReplaceCopyFromReg_128(N, DAG, Results); + return; } } diff --git a/llvm/lib/Target/NVPTX/NVPTXISelLowering.h b/llvm/lib/Target/NVPTX/NVPTXISelLowering.h index e211286..6326296 100644 --- a/llvm/lib/Target/NVPTX/NVPTXISelLowering.h +++ b/llvm/lib/Target/NVPTX/NVPTXISelLowering.h @@ -640,6 +640,14 @@ private: SDValue LowerVAARG(SDValue Op, SelectionDAG &DAG) const; SDValue LowerVASTART(SDValue Op, SelectionDAG &DAG) const; + SDValue LowerCopyToReg_128(SDValue Op, SelectionDAG &DAG) const; + unsigned getNumRegisters(LLVMContext &Context, EVT VT, + std::optional<MVT> RegisterVT) const override; + bool + splitValueIntoRegisterParts(SelectionDAG &DAG, const SDLoc &DL, SDValue Val, + SDValue *Parts, unsigned NumParts, MVT PartVT, + std::optional<CallingConv::ID> CC) const override; + void ReplaceNodeResults(SDNode *N, SmallVectorImpl<SDValue> &Results, SelectionDAG &DAG) const override; SDValue PerformDAGCombine(SDNode *N, DAGCombinerInfo &DCI) const override; diff --git a/llvm/lib/Target/NVPTX/NVPTXInstrInfo.cpp b/llvm/lib/Target/NVPTX/NVPTXInstrInfo.cpp index b0d792b..673858f 100644 --- a/llvm/lib/Target/NVPTX/NVPTXInstrInfo.cpp +++ b/llvm/lib/Target/NVPTX/NVPTXInstrInfo.cpp @@ -51,6 +51,8 @@ void NVPTXInstrInfo::copyPhysReg(MachineBasicBlock &MBB, } else if (DestRC == &NVPTX::Int64RegsRegClass) { Op = (SrcRC == &NVPTX::Int64RegsRegClass ? NVPTX::IMOV64rr : NVPTX::BITCONVERT_64_F2I); + } else if (DestRC == &NVPTX::Int128RegsRegClass) { + Op = NVPTX::IMOV128rr; } else if (DestRC == &NVPTX::Float32RegsRegClass) { Op = (SrcRC == &NVPTX::Float32RegsRegClass ? NVPTX::FMOV32rr : NVPTX::BITCONVERT_32_I2F); diff --git a/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td b/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td index c4c35a1..827febe 100644 --- a/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td +++ b/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td @@ -2097,6 +2097,8 @@ let IsSimpleMove=1, hasSideEffects=0 in { "mov.u32 \t$dst, $sss;", []>; def IMOV64rr : NVPTXInst<(outs Int64Regs:$dst), (ins Int64Regs:$sss), "mov.u64 \t$dst, $sss;", []>; + def IMOV128rr : NVPTXInst<(outs Int128Regs:$dst), (ins Int128Regs:$sss), + "mov.b128 \t$dst, $sss;", []>; def IMOVB16rr : NVPTXInst<(outs Int16Regs:$dst), (ins Int16Regs:$sss), "mov.b16 \t$dst, $sss;", []>; @@ -3545,6 +3547,9 @@ let hasSideEffects = false in { def V2I32toI64 : NVPTXInst<(outs Int64Regs:$d), (ins Int32Regs:$s1, Int32Regs:$s2), "mov.b64 \t$d, {{$s1, $s2}};", []>; + def V2I64toI128 : NVPTXInst<(outs Int128Regs:$d), + (ins Int64Regs:$s1, Int64Regs:$s2), + "mov.b128 \t$d, {{$s1, $s2}};", []>; def V2F32toF64 : NVPTXInst<(outs Float64Regs:$d), (ins Float32Regs:$s1, Float32Regs:$s2), "mov.b64 \t$d, {{$s1, $s2}};", []>; @@ -3560,6 +3565,9 @@ let hasSideEffects = false in { def I64toV2I32 : NVPTXInst<(outs Int32Regs:$d1, Int32Regs:$d2), (ins Int64Regs:$s), "mov.b64 \t{{$d1, $d2}}, $s;", []>; + def I128toV2I64: NVPTXInst<(outs Int64Regs:$d1, Int64Regs:$d2), + (ins Int128Regs:$s), + "mov.b128 \t{{$d1, $d2}}, $s;", []>; def F64toV2F32 : NVPTXInst<(outs Float32Regs:$d1, Float32Regs:$d2), (ins Float64Regs:$s), "mov.b64 \t{{$d1, $d2}}, $s;", []>; @@ -3629,7 +3637,7 @@ def : Pat<(i32 (ctlz (i32 Int32Regs:$a))), (CLZr32 Int32Regs:$a)>; // ptx value to 64 bits to match the ISD node's semantics, unless we know we're // truncating back down to 32 bits. def : Pat<(i64 (ctlz Int64Regs:$a)), (CVT_u64_u32 (CLZr64 Int64Regs:$a), CvtNONE)>; -def : Pat<(i32 (trunc (ctlz Int64Regs:$a))), (CLZr64 Int64Regs:$a)>; +def : Pat<(i32 (trunc (i64 (ctlz Int64Regs:$a)))), (CLZr64 Int64Regs:$a)>; // For 16-bit ctlz, we zero-extend to 32-bit, perform the count, then trunc the // result back to 16-bits if necessary. We also need to subtract 16 because @@ -3667,7 +3675,7 @@ def : Pat<(i32 (ctpop (i32 Int32Regs:$a))), (POPCr32 Int32Regs:$a)>; // pattern that avoids the type conversion if we're truncating the result to // i32 anyway. def : Pat<(ctpop Int64Regs:$a), (CVT_u64_u32 (POPCr64 Int64Regs:$a), CvtNONE)>; -def : Pat<(i32 (trunc (ctpop Int64Regs:$a))), (POPCr64 Int64Regs:$a)>; +def : Pat<(i32 (trunc (i64 (ctpop Int64Regs:$a)))), (POPCr64 Int64Regs:$a)>; // For 16-bit, we zero-extend to 32-bit, then trunc the result back to 16-bits. // If we know that we're storing into an i32, we can avoid the final trunc. diff --git a/llvm/lib/Target/NVPTX/NVPTXMCExpr.h b/llvm/lib/Target/NVPTX/NVPTXMCExpr.h index ef99def..d95b61c 100644 --- a/llvm/lib/Target/NVPTX/NVPTXMCExpr.h +++ b/llvm/lib/Target/NVPTX/NVPTXMCExpr.h @@ -74,8 +74,7 @@ public: /// @} void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override { return false; } @@ -116,8 +115,7 @@ public: /// @} void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override { return false; } diff --git a/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.cpp b/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.cpp index f1213f0..a8a23f0 100644 --- a/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.cpp +++ b/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.cpp @@ -31,6 +31,8 @@ std::string getNVPTXRegClassName(TargetRegisterClass const *RC) { return ".f32"; if (RC == &NVPTX::Float64RegsRegClass) return ".f64"; + if (RC == &NVPTX::Int128RegsRegClass) + return ".b128"; if (RC == &NVPTX::Int64RegsRegClass) // We use untyped (.b) integer registers here as NVCC does. // Correctness of generated code does not depend on register type, @@ -67,6 +69,8 @@ std::string getNVPTXRegClassStr(TargetRegisterClass const *RC) { return "%f"; if (RC == &NVPTX::Float64RegsRegClass) return "%fd"; + if (RC == &NVPTX::Int128RegsRegClass) + return "%rq"; if (RC == &NVPTX::Int64RegsRegClass) return "%rd"; if (RC == &NVPTX::Int32RegsRegClass) diff --git a/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.td b/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.td index b5231a9..2011f0f 100644 --- a/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.td +++ b/llvm/lib/Target/NVPTX/NVPTXRegisterInfo.td @@ -37,6 +37,7 @@ foreach i = 0...4 in { def RS#i : NVPTXReg<"%rs"#i>; // 16-bit def R#i : NVPTXReg<"%r"#i>; // 32-bit def RL#i : NVPTXReg<"%rd"#i>; // 64-bit + def RQ#i : NVPTXReg<"%rq"#i>; // 128-bit def H#i : NVPTXReg<"%h"#i>; // 16-bit float def HH#i : NVPTXReg<"%hh"#i>; // 2x16-bit float def F#i : NVPTXReg<"%f"#i>; // 32-bit float @@ -62,6 +63,8 @@ def Int32Regs : NVPTXRegClass<[i32, v2f16, v2bf16, v2i16, v4i8], 32, (add (sequence "R%u", 0, 4), VRFrame32, VRFrameLocal32)>; def Int64Regs : NVPTXRegClass<[i64], 64, (add (sequence "RL%u", 0, 4), VRFrame64, VRFrameLocal64)>; +// 128-bit regs are not defined as general regs in NVPTX. They are used for inlineASM only. +def Int128Regs : NVPTXRegClass<[i128], 128, (add (sequence "RQ%u", 0, 4))>; def Float32Regs : NVPTXRegClass<[f32], 32, (add (sequence "F%u", 0, 4))>; def Float64Regs : NVPTXRegClass<[f64], 64, (add (sequence "FL%u", 0, 4))>; def Int32ArgRegs : NVPTXRegClass<[i32], 32, (add (sequence "ia%u", 0, 4))>; diff --git a/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.cpp b/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.cpp index 6cd04ee..cc1d981 100644 --- a/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.cpp +++ b/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.cpp @@ -99,13 +99,11 @@ PPCMCExpr::evaluateAsInt64(int64_t Value) const { llvm_unreachable("Invalid kind!"); } -bool -PPCMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, - const MCFixup *Fixup) const { +bool PPCMCExpr::evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, + const MCFixup *Fixup) const { MCValue Value; - if (!getSubExpr()->evaluateAsRelocatable(Value, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Value, Asm, Fixup)) return false; if (Value.isAbsolute()) { @@ -124,10 +122,10 @@ PPCMCExpr::evaluateAsRelocatableImpl(MCValue &Res, Res = MCValue::get(Result); } else { - if (!Layout) + if (!Asm || !Asm->hasLayout()) return false; - MCContext &Context = Layout->getAssembler().getContext(); + MCContext &Context = Asm->getContext(); const MCSymbolRefExpr *Sym = Value.getSymA(); MCSymbolRefExpr::VariantKind Modifier = Sym->getKind(); if (Modifier != MCSymbolRefExpr::VK_None) diff --git a/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.h b/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.h index 1dbc7ea..59ca514 100644 --- a/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.h +++ b/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCExpr.h @@ -9,7 +9,6 @@ #ifndef LLVM_LIB_TARGET_POWERPC_MCTARGETDESC_PPCMCEXPR_H #define LLVM_LIB_TARGET_POWERPC_MCTARGETDESC_PPCMCEXPR_H -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCExpr.h" #include "llvm/MC/MCValue.h" @@ -71,8 +70,7 @@ public: /// @} void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override { diff --git a/llvm/lib/Target/RISCV/MCTargetDesc/RISCVAsmBackend.cpp b/llvm/lib/Target/RISCV/MCTargetDesc/RISCVAsmBackend.cpp index 445e589..cb64b64 100644 --- a/llvm/lib/Target/RISCV/MCTargetDesc/RISCVAsmBackend.cpp +++ b/llvm/lib/Target/RISCV/MCTargetDesc/RISCVAsmBackend.cpp @@ -10,7 +10,6 @@ #include "RISCVMCExpr.h" #include "llvm/ADT/APInt.h" #include "llvm/MC/MCAsmInfo.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCDirectives.h" @@ -215,7 +214,7 @@ bool RISCVAsmBackend::relaxDwarfLineAddr(const MCAssembler &Asm, int64_t Value; [[maybe_unused]] bool IsAbsolute = - AddrDelta.evaluateKnownAbsolute(Value, *Asm.getLayout()); + AddrDelta.evaluateKnownAbsolute(Value, Asm); assert(IsAbsolute && "CFA with invalid expression"); Data.clear(); @@ -271,17 +270,16 @@ bool RISCVAsmBackend::relaxDwarfLineAddr(const MCAssembler &Asm, bool RISCVAsmBackend::relaxDwarfCFA(const MCAssembler &Asm, MCDwarfCallFrameFragment &DF, bool &WasRelaxed) const { - auto &Layout = *Asm.getLayout(); const MCExpr &AddrDelta = DF.getAddrDelta(); SmallVectorImpl<char> &Data = DF.getContents(); SmallVectorImpl<MCFixup> &Fixups = DF.getFixups(); size_t OldSize = Data.size(); int64_t Value; - if (AddrDelta.evaluateAsAbsolute(Value, Layout)) + if (AddrDelta.evaluateAsAbsolute(Value, Asm)) return false; [[maybe_unused]] bool IsAbsolute = - AddrDelta.evaluateKnownAbsolute(Value, Layout); + AddrDelta.evaluateKnownAbsolute(Value, Asm); assert(IsAbsolute && "CFA with invalid expression"); Data.clear(); @@ -341,8 +339,7 @@ std::pair<bool, bool> RISCVAsmBackend::relaxLEB128(const MCAssembler &Asm, LF.getFixups().push_back( MCFixup::create(0, &Expr, FK_Data_leb128, Expr.getLoc())); } - return std::make_pair(Expr.evaluateKnownAbsolute(Value, *Asm.getLayout()), - false); + return std::make_pair(Expr.evaluateKnownAbsolute(Value, Asm), false); } // Given a compressed control flow instruction this function returns @@ -549,8 +546,7 @@ bool RISCVAsmBackend::evaluateTargetFixup(const MCAssembler &Asm, // MCAssembler::evaluateFixup will emit an error for this case when it sees // the %pcrel_hi, so don't duplicate it when also seeing the %pcrel_lo. const MCExpr *AUIPCExpr = AUIPCFixup->getValue(); - if (!AUIPCExpr->evaluateAsRelocatable(AUIPCTarget, Asm.getLayout(), - AUIPCFixup)) + if (!AUIPCExpr->evaluateAsRelocatable(AUIPCTarget, &Asm, AUIPCFixup)) return true; break; } diff --git a/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.cpp b/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.cpp index d83dadd..16d0796 100644 --- a/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.cpp +++ b/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.cpp @@ -15,7 +15,6 @@ #include "MCTargetDesc/RISCVAsmBackend.h" #include "RISCVFixupKinds.h" #include "llvm/BinaryFormat/ELF.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCStreamer.h" @@ -90,7 +89,7 @@ const MCFixup *RISCVMCExpr::getPCRelHiFixup(const MCFragment **DFOut) const { } bool RISCVMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { // Explicitly drop the layout and assembler to prevent any symbolic folding in // the expression handling. This is required to preserve symbolic difference diff --git a/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.h b/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.h index fcc4c5c..d144548 100644 --- a/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.h +++ b/llvm/lib/Target/RISCV/MCTargetDesc/RISCVMCExpr.h @@ -69,7 +69,7 @@ public: const MCFixup *getPCRelHiFixup(const MCFragment **DFOut) const; void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override { diff --git a/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp b/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp index bf69334..7e6eef4 100644 --- a/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp +++ b/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp @@ -586,13 +586,14 @@ public: } // Most AVLIsReg infos will have a single defining MachineInstr, unless it was // a PHI node. In that case getAVLVNInfo()->def will point to the block - // boundary slot. If LiveIntervals isn't available, then nullptr is returned. + // boundary slot and this will return nullptr. If LiveIntervals isn't + // available, nullptr is also returned. const MachineInstr *getAVLDefMI(const LiveIntervals *LIS) const { assert(hasAVLReg()); - if (!LIS) + if (!LIS || getAVLVNInfo()->isPHIDef()) return nullptr; auto *MI = LIS->getInstructionFromIndex(getAVLVNInfo()->def); - assert(!(getAVLVNInfo()->isPHIDef() && MI)); + assert(MI); return MI; } diff --git a/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp b/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp index 35d0b34..f76aef7 100644 --- a/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp +++ b/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp @@ -497,9 +497,6 @@ void RISCVPassConfig::addPreSched2() { } void RISCVPassConfig::addPreEmitPass() { - addPass(&BranchRelaxationPassID); - addPass(createRISCVMakeCompressibleOptPass()); - // TODO: It would potentially be better to schedule copy propagation after // expanding pseudos (in addPreEmitPass2). However, performing copy // propagation after the machine outliner (which runs after addPreEmitPass) @@ -508,6 +505,8 @@ void RISCVPassConfig::addPreEmitPass() { if (TM->getOptLevel() >= CodeGenOptLevel::Default && EnableRISCVCopyPropagation) addPass(createMachineCopyPropagationPass(true)); + addPass(&BranchRelaxationPassID); + addPass(createRISCVMakeCompressibleOptPass()); } void RISCVPassConfig::addPreEmitPass2() { diff --git a/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.cpp b/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.cpp index 4688837..3da19a8 100644 --- a/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.cpp +++ b/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.cpp @@ -180,11 +180,10 @@ Sparc::Fixups SparcMCExpr::getFixupKind(SparcMCExpr::VariantKind Kind) { } } -bool -SparcMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, - const MCFixup *Fixup) const { - return getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup); +bool SparcMCExpr::evaluateAsRelocatableImpl(MCValue &Res, + const MCAssembler *Asm, + const MCFixup *Fixup) const { + return getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup); } static void fixELFSymbolsInTLSFixupsImpl(const MCExpr *Expr, MCAssembler &Asm) { diff --git a/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.h b/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.h index d26a748..74f90ae 100644 --- a/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.h +++ b/llvm/lib/Target/Sparc/MCTargetDesc/SparcMCExpr.h @@ -94,8 +94,7 @@ public: /// @} void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override { diff --git a/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.cpp b/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.cpp index 647cf76..4fa2028 100644 --- a/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.cpp +++ b/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.cpp @@ -37,9 +37,9 @@ void SystemZMCExpr::printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const { } bool SystemZMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { - if (!getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup)) return false; Res = diff --git a/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.h b/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.h index 62e6109..98f3a23 100644 --- a/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.h +++ b/llvm/lib/Target/SystemZ/MCTargetDesc/SystemZMCExpr.h @@ -45,7 +45,7 @@ public: StringRef getVariantKindName() const; void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override { Streamer.visitUsedExpr(*getSubExpr()); diff --git a/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.cpp b/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.cpp index a104510..45c768e 100644 --- a/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.cpp +++ b/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.cpp @@ -173,10 +173,9 @@ VE::Fixups VEMCExpr::getFixupKind(VEMCExpr::VariantKind Kind) { } } -bool VEMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, +bool VEMCExpr::evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const { - if (!getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup)) + if (!getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup)) return false; Res = diff --git a/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.h b/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.h index 2da956d..ec71f9d 100644 --- a/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.h +++ b/llvm/lib/Target/VE/MCTargetDesc/VEMCExpr.h @@ -69,7 +69,7 @@ public: /// @} void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override { diff --git a/llvm/lib/Target/WebAssembly/WebAssembly.td b/llvm/lib/Target/WebAssembly/WebAssembly.td index 0dd6744..9761861 100644 --- a/llvm/lib/Target/WebAssembly/WebAssembly.td +++ b/llvm/lib/Target/WebAssembly/WebAssembly.td @@ -110,13 +110,17 @@ def : ProcessorModel<"mvp", NoSchedModel, []>; // consideration given to available support in relevant engines and tools, and // the importance of the features. def : ProcessorModel<"generic", NoSchedModel, - [FeatureSignExt, FeatureMutableGlobals]>; + [FeatureMultivalue, FeatureMutableGlobals, + FeatureReferenceTypes, FeatureSignExt]>; // Latest and greatest experimental version of WebAssembly. Bugs included! def : ProcessorModel<"bleeding-edge", NoSchedModel, - [FeatureSIMD128, FeatureAtomics, - FeatureNontrappingFPToInt, FeatureSignExt, - FeatureMutableGlobals, FeatureBulkMemory, + [FeatureAtomics, FeatureBulkMemory, + FeatureExceptionHandling, FeatureExtendedConst, + FeatureHalfPrecision, FeatureMultiMemory, + FeatureMultivalue, FeatureMutableGlobals, + FeatureNontrappingFPToInt, FeatureRelaxedSIMD, + FeatureReferenceTypes, FeatureSIMD128, FeatureSignExt, FeatureTailCall]>; //===----------------------------------------------------------------------===// diff --git a/llvm/lib/Target/X86/MCTargetDesc/X86AsmBackend.cpp b/llvm/lib/Target/X86/MCTargetDesc/X86AsmBackend.cpp index 670ff35..09a6d57 100644 --- a/llvm/lib/Target/X86/MCTargetDesc/X86AsmBackend.cpp +++ b/llvm/lib/Target/X86/MCTargetDesc/X86AsmBackend.cpp @@ -13,7 +13,6 @@ #include "llvm/BinaryFormat/ELF.h" #include "llvm/BinaryFormat/MachO.h" #include "llvm/MC/MCAsmBackend.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCCodeEmitter.h" #include "llvm/MC/MCContext.h" @@ -202,7 +201,7 @@ public: bool padInstructionEncoding(MCRelaxableFragment &RF, MCCodeEmitter &Emitter, unsigned &RemainingSize) const; - void finishLayout(MCAssembler const &Asm, MCAsmLayout &Layout) const override; + void finishLayout(const MCAssembler &Asm) const override; unsigned getMaximumNopSize(const MCSubtargetInfo &STI) const override; @@ -857,8 +856,7 @@ bool X86AsmBackend::padInstructionEncoding(MCRelaxableFragment &RF, return Changed; } -void X86AsmBackend::finishLayout(MCAssembler const &Asm, - MCAsmLayout &Layout) const { +void X86AsmBackend::finishLayout(MCAssembler const &Asm) const { // See if we can further relax some instructions to cut down on the number of // nop bytes required for code alignment. The actual win is in reducing // instruction count, not number of bytes. Modern X86-64 can easily end up diff --git a/llvm/lib/Target/X86/MCTargetDesc/X86MCExpr.h b/llvm/lib/Target/X86/MCTargetDesc/X86MCExpr.h index db91d38..c159d30 100644 --- a/llvm/lib/Target/X86/MCTargetDesc/X86MCExpr.h +++ b/llvm/lib/Target/X86/MCTargetDesc/X86MCExpr.h @@ -53,7 +53,7 @@ public: OS << X86ATTInstPrinter::getRegisterName(RegNo); } - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override { return false; } diff --git a/llvm/lib/Target/X86/MCTargetDesc/X86MachObjectWriter.cpp b/llvm/lib/Target/X86/MCTargetDesc/X86MachObjectWriter.cpp index d28195f..ec95b1f 100644 --- a/llvm/lib/Target/X86/MCTargetDesc/X86MachObjectWriter.cpp +++ b/llvm/lib/Target/X86/MCTargetDesc/X86MachObjectWriter.cpp @@ -12,7 +12,6 @@ #include "llvm/BinaryFormat/MachO.h" #include "llvm/MC/MCAsmInfo.h" #include "llvm/MC/MCAsmInfoDarwin.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCMachObjectWriter.h" @@ -185,12 +184,10 @@ void X86MachObjectWriter::RecordX86_64Relocation( return; } - Value += - Writer->getSymbolAddress(*A, *Asm.getLayout()) - - (!A_Base ? 0 : Writer->getSymbolAddress(*A_Base, *Asm.getLayout())); - Value -= - Writer->getSymbolAddress(*B, *Asm.getLayout()) - - (!B_Base ? 0 : Writer->getSymbolAddress(*B_Base, *Asm.getLayout())); + Value += Writer->getSymbolAddress(*A, Asm) - + (!A_Base ? 0 : Writer->getSymbolAddress(*A_Base, Asm)); + Value -= Writer->getSymbolAddress(*B, Asm) - + (!B_Base ? 0 : Writer->getSymbolAddress(*B_Base, Asm)); if (!A_Base) Index = A->getFragment()->getParent()->getOrdinal() + 1; @@ -237,7 +234,7 @@ void X86MachObjectWriter::RecordX86_64Relocation( } else if (Symbol->isInSection() && !Symbol->isVariable()) { // The index is the section ordinal (1-based). Index = Symbol->getFragment()->getParent()->getOrdinal() + 1; - Value += Writer->getSymbolAddress(*Symbol, *Asm.getLayout()); + Value += Writer->getSymbolAddress(*Symbol, Asm); if (IsPCRel) Value -= FixupAddress + (1 << Log2Size); @@ -376,7 +373,7 @@ bool X86MachObjectWriter::recordScatteredRelocation(MachObjectWriter *Writer, return false; } - uint32_t Value = Writer->getSymbolAddress(*A, *Asm.getLayout()); + uint32_t Value = Writer->getSymbolAddress(*A, Asm); uint64_t SecAddr = Writer->getSectionAddress(A->getFragment()->getParent()); FixedValue += SecAddr; uint32_t Value2 = 0; @@ -399,7 +396,7 @@ bool X86MachObjectWriter::recordScatteredRelocation(MachObjectWriter *Writer, // pedantic compatibility with 'as'. Type = A->isExternal() ? (unsigned)MachO::GENERIC_RELOC_SECTDIFF : (unsigned)MachO::GENERIC_RELOC_LOCAL_SECTDIFF; - Value2 = Writer->getSymbolAddress(*SB, *Asm.getLayout()); + Value2 = Writer->getSymbolAddress(*SB, Asm); FixedValue -= Writer->getSectionAddress(SB->getFragment()->getParent()); } @@ -476,7 +473,7 @@ void X86MachObjectWriter::recordTLVPRelocation(MachObjectWriter *Writer, Writer->getFragmentAddress(Asm, Fragment) + Fixup.getOffset(); IsPCRel = 1; FixedValue = FixupAddress - - Writer->getSymbolAddress(SymB->getSymbol(), *Asm.getLayout()) + + Writer->getSymbolAddress(SymB->getSymbol(), Asm) + Target.getConstant(); FixedValue += 1ULL << Log2Size; } else { diff --git a/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.cpp b/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.cpp index cafd8b7..900ce3c 100644 --- a/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.cpp +++ b/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.cpp @@ -41,9 +41,9 @@ void XtensaMCExpr::printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const { } bool XtensaMCExpr::evaluateAsRelocatableImpl(MCValue &Res, - const MCAsmLayout *Layout, + const MCAssembler *Asm, const MCFixup *Fixup) const { - return getSubExpr()->evaluateAsRelocatable(Res, Layout, Fixup); + return getSubExpr()->evaluateAsRelocatable(Res, Asm, Fixup); } void XtensaMCExpr::visitUsedExpr(MCStreamer &Streamer) const { diff --git a/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.h b/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.h index 4cc78fe..92ffb5c 100644 --- a/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.h +++ b/llvm/lib/Target/Xtensa/MCTargetDesc/XtensaMCExpr.h @@ -40,7 +40,7 @@ public: const MCExpr *getSubExpr() const { return Expr; } void printImpl(raw_ostream &OS, const MCAsmInfo *MAI) const override; - bool evaluateAsRelocatableImpl(MCValue &Res, const MCAsmLayout *Layout, + bool evaluateAsRelocatableImpl(MCValue &Res, const MCAssembler *Asm, const MCFixup *Fixup) const override; void visitUsedExpr(MCStreamer &Streamer) const override; MCFragment *findAssociatedFragment() const override { diff --git a/llvm/lib/TargetParser/RISCVISAInfo.cpp b/llvm/lib/TargetParser/RISCVISAInfo.cpp index 2a6ecee..869be57 100644 --- a/llvm/lib/TargetParser/RISCVISAInfo.cpp +++ b/llvm/lib/TargetParser/RISCVISAInfo.cpp @@ -171,11 +171,6 @@ findDefaultVersion(StringRef ExtName) { return std::nullopt; } -bool RISCVISAInfo::addExtension(StringRef ExtName, - RISCVISAUtils::ExtensionVersion Version) { - return Exts.emplace(ExtName, Version).second; -} - static StringRef getExtensionTypeDesc(StringRef Ext) { if (Ext.starts_with('s')) return "standard supervisor-level extension"; @@ -431,7 +426,7 @@ RISCVISAInfo::parseFeatures(unsigned XLen, continue; if (Add) - ISAInfo->addExtension(ExtName, ExtensionInfoIterator->Version); + ISAInfo->Exts[ExtName.str()] = ExtensionInfoIterator->Version; else ISAInfo->Exts.erase(ExtName.str()); } @@ -513,7 +508,11 @@ RISCVISAInfo::parseNormalizedArchString(StringRef Arch) { "'" + Twine(ExtName[0]) + "' must be followed by a letter"); - if (!ISAInfo->addExtension(ExtName, {MajorVersion, MinorVersion})) + if (!ISAInfo->Exts + .emplace( + ExtName.str(), + RISCVISAUtils::ExtensionVersion{MajorVersion, MinorVersion}) + .second) return createStringError(errc::invalid_argument, "duplicate extension '" + ExtName + "'"); } @@ -586,6 +585,10 @@ RISCVISAInfo::parseArchString(StringRef Arch, bool EnableExperimentalExtension, // The canonical order specified in ISA manual. // Ref: Table 22.1 in RISC-V User-Level ISA V2.2 char Baseline = Arch.front(); + // Skip the baseline. + StringRef Exts = Arch.drop_front(); + + unsigned Major, Minor, ConsumeLength; // First letter should be 'e', 'i' or 'g'. switch (Baseline) { @@ -595,20 +598,29 @@ RISCVISAInfo::parseArchString(StringRef Arch, bool EnableExperimentalExtension, "\' should be 'e', 'i' or 'g'"); case 'e': case 'i': + // Baseline is `i` or `e` + if (auto E = getExtensionVersion( + StringRef(&Baseline, 1), Exts, Major, Minor, ConsumeLength, + EnableExperimentalExtension, ExperimentalExtensionVersionCheck)) { + if (!IgnoreUnknown) + return std::move(E); + // If IgnoreUnknown, then ignore an unrecognised version of the baseline + // ISA and just use the default supported version. + consumeError(std::move(E)); + auto Version = findDefaultVersion(StringRef(&Baseline, 1)); + Major = Version->Major; + Minor = Version->Minor; + } + + // Postpone AddExtension until end of this function + SeenExtMap[StringRef(&Baseline, 1).str()] = {Major, Minor}; break; case 'g': // g expands to extensions in RISCVGImplications. if (Arch.size() > 1 && isDigit(Arch[1])) return createStringError(errc::invalid_argument, "version not supported for 'g'"); - break; - } - // Skip baseline. - StringRef Exts = Arch.drop_front(1); - - unsigned Major, Minor, ConsumeLength; - if (Baseline == 'g') { // Versions for g are disallowed, and this was checked for previously. ConsumeLength = 0; @@ -621,23 +633,7 @@ RISCVISAInfo::parseArchString(StringRef Arch, bool EnableExperimentalExtension, // Postpone AddExtension until end of this function SeenExtMap[Ext] = {Version->Major, Version->Minor}; } - } else { - // Baseline is `i` or `e` - if (auto E = getExtensionVersion( - StringRef(&Baseline, 1), Exts, Major, Minor, ConsumeLength, - EnableExperimentalExtension, ExperimentalExtensionVersionCheck)) { - if (!IgnoreUnknown) - return std::move(E); - // If IgnoreUnknown, then ignore an unrecognised version of the baseline - // ISA and just use the default supported version. - consumeError(std::move(E)); - auto Version = findDefaultVersion(StringRef(&Baseline, 1)); - Major = Version->Major; - Minor = Version->Minor; - } - - // Postpone AddExtension until end of this function - SeenExtMap[StringRef(&Baseline, 1).str()] = {Major, Minor}; + break; } // Consume the base ISA version number and any '_' between rvxxx and the @@ -695,13 +691,13 @@ RISCVISAInfo::parseArchString(StringRef Arch, bool EnableExperimentalExtension, if (auto E = getExtensionVersion(Name, Vers, Major, Minor, ConsumeLength, EnableExperimentalExtension, ExperimentalExtensionVersionCheck)) { - if (IgnoreUnknown) { - consumeError(std::move(E)); - if (Name.size() == 1) - Ext = Ext.substr(ConsumeLength); - continue; - } - return E; + if (!IgnoreUnknown) + return E; + + consumeError(std::move(E)); + if (Name.size() == 1) + Ext = Ext.substr(ConsumeLength); + continue; } if (Name.size() == 1) @@ -722,11 +718,10 @@ RISCVISAInfo::parseArchString(StringRef Arch, bool EnableExperimentalExtension, // Check all Extensions are supported. for (auto &SeenExtAndVers : SeenExtMap) { const std::string &ExtName = SeenExtAndVers.first; - RISCVISAUtils::ExtensionVersion ExtVers = SeenExtAndVers.second; if (!RISCVISAInfo::isSupportedExtension(ExtName)) return getStringErrorForInvalidExt(ExtName); - ISAInfo->addExtension(ExtName, ExtVers); + ISAInfo->Exts[ExtName] = SeenExtAndVers.second; } return RISCVISAInfo::postProcessAndChecking(std::move(ISAInfo)); @@ -830,7 +825,7 @@ void RISCVISAInfo::updateImplication() { // implied if (!HasE && !HasI) { auto Version = findDefaultVersion("i"); - addExtension("i", *Version); + Exts["i"] = *Version; } if (HasE && HasI) @@ -854,7 +849,7 @@ void RISCVISAInfo::updateImplication() { if (Exts.count(ImpliedExt)) return; auto Version = findDefaultVersion(ImpliedExt); - addExtension(ImpliedExt, *Version); + Exts[ImpliedExt] = *Version; WorkList.push_back(ImpliedExt); }); } @@ -863,7 +858,7 @@ void RISCVISAInfo::updateImplication() { if (XLen == 32 && Exts.count("zce") && Exts.count("f") && !Exts.count("zcf")) { auto Version = findDefaultVersion("zcf"); - addExtension("zcf", *Version); + Exts["zcf"] = *Version; } } @@ -890,7 +885,7 @@ void RISCVISAInfo::updateCombination() { }); if (HasAllRequiredFeatures) { auto Version = findDefaultVersion(CombineExt); - addExtension(CombineExt, *Version); + Exts[CombineExt.str()] = *Version; MadeChange = true; } } diff --git a/llvm/lib/Transforms/IPO/AlwaysInliner.cpp b/llvm/lib/Transforms/IPO/AlwaysInliner.cpp index cc375f9..1f787c7 100644 --- a/llvm/lib/Transforms/IPO/AlwaysInliner.cpp +++ b/llvm/lib/Transforms/IPO/AlwaysInliner.cpp @@ -15,12 +15,12 @@ #include "llvm/ADT/SetVector.h" #include "llvm/Analysis/AliasAnalysis.h" #include "llvm/Analysis/AssumptionCache.h" +#include "llvm/Analysis/InlineAdvisor.h" #include "llvm/Analysis/InlineCost.h" #include "llvm/Analysis/OptimizationRemarkEmitter.h" #include "llvm/Analysis/ProfileSummaryInfo.h" #include "llvm/IR/Module.h" #include "llvm/InitializePasses.h" -#include "llvm/Transforms/IPO/Inliner.h" #include "llvm/Transforms/Utils/Cloning.h" #include "llvm/Transforms/Utils/ModuleUtils.h" @@ -37,86 +37,73 @@ bool AlwaysInlineImpl( function_ref<BlockFrequencyInfo &(Function &)> GetBFI) { SmallSetVector<CallBase *, 16> Calls; bool Changed = false; - SmallVector<Function *, 16> InlinedFunctions; - for (Function &F : M) { - // When callee coroutine function is inlined into caller coroutine function - // before coro-split pass, - // coro-early pass can not handle this quiet well. - // So we won't inline the coroutine function if it have not been unsplited + SmallVector<Function *, 16> InlinedComdatFunctions; + + for (Function &F : make_early_inc_range(M)) { if (F.isPresplitCoroutine()) continue; - if (!F.isDeclaration() && isInlineViable(F).isSuccess()) { - Calls.clear(); - - for (User *U : F.users()) - if (auto *CB = dyn_cast<CallBase>(U)) - if (CB->getCalledFunction() == &F && - CB->hasFnAttr(Attribute::AlwaysInline) && - !CB->getAttributes().hasFnAttr(Attribute::NoInline)) - Calls.insert(CB); - - for (CallBase *CB : Calls) { - Function *Caller = CB->getCaller(); - OptimizationRemarkEmitter ORE(Caller); - DebugLoc DLoc = CB->getDebugLoc(); - BasicBlock *Block = CB->getParent(); - - InlineFunctionInfo IFI(GetAssumptionCache, &PSI, - GetBFI ? &GetBFI(*Caller) : nullptr, - GetBFI ? &GetBFI(F) : nullptr); - - InlineResult Res = InlineFunction(*CB, IFI, /*MergeAttributes=*/true, - &GetAAR(F), InsertLifetime); - if (!Res.isSuccess()) { - ORE.emit([&]() { - return OptimizationRemarkMissed(DEBUG_TYPE, "NotInlined", DLoc, - Block) - << "'" << ore::NV("Callee", &F) << "' is not inlined into '" - << ore::NV("Caller", Caller) - << "': " << ore::NV("Reason", Res.getFailureReason()); - }); - continue; - } - - emitInlinedIntoBasedOnCost( - ORE, DLoc, Block, F, *Caller, - InlineCost::getAlways("always inline attribute"), - /*ForProfileContext=*/false, DEBUG_TYPE); + if (F.isDeclaration() || !isInlineViable(F).isSuccess()) + continue; - Changed = true; + Calls.clear(); + + for (User *U : F.users()) + if (auto *CB = dyn_cast<CallBase>(U)) + if (CB->getCalledFunction() == &F && + CB->hasFnAttr(Attribute::AlwaysInline) && + !CB->getAttributes().hasFnAttr(Attribute::NoInline)) + Calls.insert(CB); + + for (CallBase *CB : Calls) { + Function *Caller = CB->getCaller(); + OptimizationRemarkEmitter ORE(Caller); + DebugLoc DLoc = CB->getDebugLoc(); + BasicBlock *Block = CB->getParent(); + + InlineFunctionInfo IFI(GetAssumptionCache, &PSI, + GetBFI ? &GetBFI(*Caller) : nullptr, + GetBFI ? &GetBFI(F) : nullptr); + + InlineResult Res = InlineFunction(*CB, IFI, /*MergeAttributes=*/true, + &GetAAR(F), InsertLifetime); + if (!Res.isSuccess()) { + ORE.emit([&]() { + return OptimizationRemarkMissed(DEBUG_TYPE, "NotInlined", DLoc, Block) + << "'" << ore::NV("Callee", &F) << "' is not inlined into '" + << ore::NV("Caller", Caller) + << "': " << ore::NV("Reason", Res.getFailureReason()); + }); + continue; } - if (F.hasFnAttribute(Attribute::AlwaysInline)) { - // Remember to try and delete this function afterward. This both avoids - // re-walking the rest of the module and avoids dealing with any - // iterator invalidation issues while deleting functions. - InlinedFunctions.push_back(&F); - } + emitInlinedIntoBasedOnCost( + ORE, DLoc, Block, F, *Caller, + InlineCost::getAlways("always inline attribute"), + /*ForProfileContext=*/false, DEBUG_TYPE); + + Changed = true; } - } - // Remove any live functions. - erase_if(InlinedFunctions, [&](Function *F) { - F->removeDeadConstantUsers(); - return !F->isDefTriviallyDead(); - }); - - // Delete the non-comdat ones from the module and also from our vector. - auto NonComdatBegin = partition( - InlinedFunctions, [&](Function *F) { return F->hasComdat(); }); - for (Function *F : make_range(NonComdatBegin, InlinedFunctions.end())) { - M.getFunctionList().erase(F); - Changed = true; + F.removeDeadConstantUsers(); + if (F.hasFnAttribute(Attribute::AlwaysInline) && F.isDefTriviallyDead()) { + // Remember to try and delete this function afterward. This allows to call + // filterDeadComdatFunctions() only once. + if (F.hasComdat()) { + InlinedComdatFunctions.push_back(&F); + } else { + M.getFunctionList().erase(F); + Changed = true; + } + } } - InlinedFunctions.erase(NonComdatBegin, InlinedFunctions.end()); - if (!InlinedFunctions.empty()) { + if (!InlinedComdatFunctions.empty()) { // Now we just have the comdat functions. Filter out the ones whose comdats // are not actually dead. - filterDeadComdatFunctions(InlinedFunctions); + filterDeadComdatFunctions(InlinedComdatFunctions); // The remaining functions are actually dead. - for (Function *F : InlinedFunctions) { + for (Function *F : InlinedComdatFunctions) { M.getFunctionList().erase(F); Changed = true; } diff --git a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp index a47f976..2816a85 100644 --- a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp +++ b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp @@ -1621,13 +1621,6 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) { return true; }; - const auto *F = getAnchorScope(); - const auto *CI = - F ? A.getInfoCache().getAnalysisResultForFunction<CycleAnalysis>(*F) - : nullptr; - const auto *TLI = - F ? A.getInfoCache().getTargetLibraryInfoForFunction(*F) : nullptr; - auto UsePred = [&](const Use &U, bool &Follow) -> bool { Value *CurPtr = U.get(); User *Usr = U.getUser(); @@ -1671,18 +1664,18 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) { // For PHIs we need to take care of the recurrence explicitly as the value // might change while we iterate through a loop. For now, we give up if // the PHI is not invariant. - if (isa<PHINode>(Usr)) { + if (auto *PHI = dyn_cast<PHINode>(Usr)) { // Note the order here, the Usr access might change the map, CurPtr is // already in it though. - bool IsFirstPHIUser = !OffsetInfoMap.count(Usr); - auto &UsrOI = OffsetInfoMap[Usr]; + bool IsFirstPHIUser = !OffsetInfoMap.count(PHI); + auto &UsrOI = OffsetInfoMap[PHI]; auto &PtrOI = OffsetInfoMap[CurPtr]; // Check if the PHI operand has already an unknown offset as we can't // improve on that anymore. if (PtrOI.isUnknown()) { LLVM_DEBUG(dbgs() << "[AAPointerInfo] PHI operand offset unknown " - << *CurPtr << " in " << *Usr << "\n"); + << *CurPtr << " in " << *PHI << "\n"); Follow = !UsrOI.isUnknown(); UsrOI.setUnknown(); return true; @@ -1705,7 +1698,8 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) { auto It = OffsetInfoMap.find(CurPtrBase); if (It == OffsetInfoMap.end()) { LLVM_DEBUG(dbgs() << "[AAPointerInfo] PHI operand is too complex " - << *CurPtr << " in " << *Usr << "\n"); + << *CurPtr << " in " << *PHI + << " (base: " << *CurPtrBase << ")\n"); UsrOI.setUnknown(); Follow = true; return true; @@ -1718,6 +1712,9 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) { // Cycles reported by CycleInfo. It is sufficient to check the PHIs in // every Cycle header; if such a node is marked unknown, this will // eventually propagate through the whole net of PHIs in the recurrence. + const auto *CI = + A.getInfoCache().getAnalysisResultForFunction<CycleAnalysis>( + *PHI->getFunction()); if (mayBeInCycle(CI, cast<Instruction>(Usr), /* HeaderOnly */ true)) { auto BaseOI = It->getSecond(); BaseOI.addToAll(Offset.getZExtValue()); @@ -1729,7 +1726,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) { LLVM_DEBUG( dbgs() << "[AAPointerInfo] PHI operand pointer offset mismatch " - << *CurPtr << " in " << *Usr << "\n"); + << *CurPtr << " in " << *PHI << "\n"); UsrOI.setUnknown(); Follow = true; return true; @@ -1882,6 +1879,8 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) { if (auto *CB = dyn_cast<CallBase>(Usr)) { if (CB->isLifetimeStartOrEnd()) return true; + const auto *TLI = + A.getInfoCache().getTargetLibraryInfoForFunction(*CB->getFunction()); if (getFreedOperand(CB, TLI) == U) return true; if (CB->isArgOperand(&U)) { diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp index 9291e6e..0ca56f0 100644 --- a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp +++ b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp @@ -2641,6 +2641,31 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) { return BinaryOperator::CreateFMulFMF(Src, Select, II); } + // ldexp(x, c ? exp : 0) -> c ? ldexp(x, exp) : x + // ldexp(x, c ? 0 : exp) -> c ? x : ldexp(x, exp) + /// + // TODO: If we cared, should insert a canonicalize for x + Value *SelectCond, *SelectLHS, *SelectRHS; + if (match(II->getArgOperand(1), + m_OneUse(m_Select(m_Value(SelectCond), m_Value(SelectLHS), + m_Value(SelectRHS))))) { + Value *NewLdexp = nullptr; + Value *Select = nullptr; + if (match(SelectRHS, m_ZeroInt())) { + NewLdexp = Builder.CreateLdexp(Src, SelectLHS); + Select = Builder.CreateSelect(SelectCond, NewLdexp, Src); + } else if (match(SelectLHS, m_ZeroInt())) { + NewLdexp = Builder.CreateLdexp(Src, SelectRHS); + Select = Builder.CreateSelect(SelectCond, Src, NewLdexp); + } + + if (NewLdexp) { + Select->takeName(II); + cast<Instruction>(NewLdexp)->copyFastMathFlags(II); + return replaceInstUsesWith(*II, Select); + } + } + break; } case Intrinsic::ptrauth_auth: diff --git a/llvm/lib/Transforms/InstCombine/InstCombineInternal.h b/llvm/lib/Transforms/InstCombine/InstCombineInternal.h index 318c455..64fbcc8 100644 --- a/llvm/lib/Transforms/InstCombine/InstCombineInternal.h +++ b/llvm/lib/Transforms/InstCombine/InstCombineInternal.h @@ -545,10 +545,11 @@ public: ConstantInt *&Less, ConstantInt *&Equal, ConstantInt *&Greater); - /// Attempts to replace V with a simpler value based on the demanded + /// Attempts to replace I with a simpler value based on the demanded /// bits. - Value *SimplifyDemandedUseBits(Value *V, APInt DemandedMask, KnownBits &Known, - unsigned Depth, const SimplifyQuery &Q); + Value *SimplifyDemandedUseBits(Instruction *I, const APInt &DemandedMask, + KnownBits &Known, unsigned Depth, + const SimplifyQuery &Q); using InstCombiner::SimplifyDemandedBits; bool SimplifyDemandedBits(Instruction *I, unsigned Op, const APInt &DemandedMask, KnownBits &Known, diff --git a/llvm/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp b/llvm/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp index 6cf2e71..98f085a 100644 --- a/llvm/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp +++ b/llvm/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp @@ -91,8 +91,38 @@ bool InstCombinerImpl::SimplifyDemandedBits(Instruction *I, unsigned OpNo, KnownBits &Known, unsigned Depth, const SimplifyQuery &Q) { Use &U = I->getOperandUse(OpNo); - Value *NewVal = SimplifyDemandedUseBits(U.get(), DemandedMask, Known, - Depth, Q); + Value *V = U.get(); + if (isa<Constant>(V)) { + llvm::computeKnownBits(V, Known, Depth, Q); + return false; + } + + Known.resetAll(); + if (DemandedMask.isZero()) { + // Not demanding any bits from V. + replaceUse(U, UndefValue::get(V->getType())); + return true; + } + + if (Depth == MaxAnalysisRecursionDepth) + return false; + + Instruction *VInst = dyn_cast<Instruction>(V); + if (!VInst) { + llvm::computeKnownBits(V, Known, Depth, Q); + return false; + } + + Value *NewVal; + if (VInst->hasOneUse()) { + // If the instruction has one use, we can directly simplify it. + NewVal = SimplifyDemandedUseBits(VInst, DemandedMask, Known, Depth, Q); + } else { + // If there are multiple uses of this instruction, then we can simplify + // VInst to some other value, but not modify the instruction. + NewVal = + SimplifyMultipleUseDemandedBits(VInst, DemandedMask, Known, Depth, Q); + } if (!NewVal) return false; if (Instruction* OpInst = dyn_cast<Instruction>(U)) salvageDebugInfo(*OpInst); @@ -124,50 +154,21 @@ bool InstCombinerImpl::SimplifyDemandedBits(Instruction *I, unsigned OpNo, /// operands based on the information about what bits are demanded. This returns /// some other non-null value if it found out that V is equal to another value /// in the context where the specified bits are demanded, but not for all users. -Value *InstCombinerImpl::SimplifyDemandedUseBits(Value *V, APInt DemandedMask, +Value *InstCombinerImpl::SimplifyDemandedUseBits(Instruction *I, + const APInt &DemandedMask, KnownBits &Known, unsigned Depth, const SimplifyQuery &Q) { - assert(V != nullptr && "Null pointer of Value???"); + assert(I != nullptr && "Null pointer of Value???"); assert(Depth <= MaxAnalysisRecursionDepth && "Limit Search Depth"); uint32_t BitWidth = DemandedMask.getBitWidth(); - Type *VTy = V->getType(); + Type *VTy = I->getType(); assert( (!VTy->isIntOrIntVectorTy() || VTy->getScalarSizeInBits() == BitWidth) && Known.getBitWidth() == BitWidth && "Value *V, DemandedMask and Known must have same BitWidth"); - if (isa<Constant>(V)) { - llvm::computeKnownBits(V, Known, Depth, Q); - return nullptr; - } - - Known.resetAll(); - if (DemandedMask.isZero()) // Not demanding any bits from V. - return UndefValue::get(VTy); - - if (Depth == MaxAnalysisRecursionDepth) - return nullptr; - - Instruction *I = dyn_cast<Instruction>(V); - if (!I) { - llvm::computeKnownBits(V, Known, Depth, Q); - return nullptr; // Only analyze instructions. - } - - // If there are multiple uses of this value and we aren't at the root, then - // we can't do any simplifications of the operands, because DemandedMask - // only reflects the bits demanded by *one* of the users. - if (Depth != 0 && !I->hasOneUse()) - return SimplifyMultipleUseDemandedBits(I, DemandedMask, Known, Depth, Q); - KnownBits LHSKnown(BitWidth), RHSKnown(BitWidth); - // If this is the root being simplified, allow it to have multiple uses, - // just set the DemandedMask to all bits so that we can try to simplify the - // operands. This allows visitTruncInst (for example) to simplify the - // operand of a trunc without duplicating all the logic below. - if (Depth == 0 && !V->hasOneUse()) - DemandedMask.setAllBits(); // Update flags after simplifying an operand based on the fact that some high // order bits are not demanded. @@ -1105,13 +1106,13 @@ Value *InstCombinerImpl::SimplifyDemandedUseBits(Value *V, APInt DemandedMask, } if (!KnownBitsComputed) - llvm::computeKnownBits(V, Known, Depth, Q); + llvm::computeKnownBits(I, Known, Depth, Q); break; } } - if (V->getType()->isPointerTy()) { - Align Alignment = V->getPointerAlignment(DL); + if (I->getType()->isPointerTy()) { + Align Alignment = I->getPointerAlignment(DL); Known.Zero.setLowBits(Log2(Alignment)); } @@ -1119,13 +1120,14 @@ Value *InstCombinerImpl::SimplifyDemandedUseBits(Value *V, APInt DemandedMask, // constant. We can't directly simplify pointers as a constant because of // pointer provenance. // TODO: We could return `(inttoptr const)` for pointers. - if (!V->getType()->isPointerTy() && DemandedMask.isSubsetOf(Known.Zero | Known.One)) + if (!I->getType()->isPointerTy() && + DemandedMask.isSubsetOf(Known.Zero | Known.One)) return Constant::getIntegerValue(VTy, Known.One); if (VerifyKnownBits) { - KnownBits ReferenceKnown = llvm::computeKnownBits(V, Depth, Q); + KnownBits ReferenceKnown = llvm::computeKnownBits(I, Depth, Q); if (Known != ReferenceKnown) { - errs() << "Mismatched known bits for " << *V << " in " + errs() << "Mismatched known bits for " << *I << " in " << I->getFunction()->getName() << "\n"; errs() << "computeKnownBits(): " << ReferenceKnown << "\n"; errs() << "SimplifyDemandedBits(): " << Known << "\n"; diff --git a/llvm/lib/Transforms/Instrumentation/IndirectCallPromotion.cpp b/llvm/lib/Transforms/Instrumentation/IndirectCallPromotion.cpp index 6a5aeeb..0d1f506 100644 --- a/llvm/lib/Transforms/Instrumentation/IndirectCallPromotion.cpp +++ b/llvm/lib/Transforms/Instrumentation/IndirectCallPromotion.cpp @@ -547,18 +547,17 @@ Instruction *IndirectCallPromoter::computeVTableInfos( for (size_t I = 0; I < Candidates.size(); I++) CalleeIndexMap[Candidates[I].TargetFunction] = I; - uint32_t ActualNumValueData = 0; uint64_t TotalVTableCount = 0; - auto VTableValueDataArray = getValueProfDataFromInst( - *VirtualCallInfo.VPtr, IPVK_VTableTarget, MaxNumVTableAnnotations, - ActualNumValueData, TotalVTableCount); - if (VTableValueDataArray.get() == nullptr) + auto VTableValueDataArray = + getValueProfDataFromInst(*VirtualCallInfo.VPtr, IPVK_VTableTarget, + MaxNumVTableAnnotations, TotalVTableCount); + if (VTableValueDataArray.empty()) return VPtr; // Compute the functions and counts from by each vtable. - for (size_t j = 0; j < ActualNumValueData; j++) { - uint64_t VTableVal = VTableValueDataArray[j].Value; - GUIDCountsMap[VTableVal] = VTableValueDataArray[j].Count; + for (const auto &V : VTableValueDataArray) { + uint64_t VTableVal = V.Value; + GUIDCountsMap[VTableVal] = V.Count; GlobalVariable *VTableVar = Symtab->getGlobalVariable(VTableVal); if (!VTableVar) { LLVM_DEBUG(dbgs() << " Cannot find vtable definition for " << VTableVal @@ -586,7 +585,7 @@ Instruction *IndirectCallPromoter::computeVTableInfos( // There shouldn't be duplicate GUIDs in one !prof metadata (except // duplicated zeros), so assign counters directly won't cause overwrite or // counter loss. - Candidate.VTableGUIDAndCounts[VTableVal] = VTableValueDataArray[j].Count; + Candidate.VTableGUIDAndCounts[VTableVal] = V.Count; Candidate.AddressPoints.push_back( getOrCreateVTableAddressPointVar(VTableVar, AddressPointOffset)); } diff --git a/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp b/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp index 88adeb5..875d3ea 100644 --- a/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp +++ b/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp @@ -85,6 +85,7 @@ STATISTIC(NumOverflows, "Number of overflow checks removed"); STATISTIC(NumSaturating, "Number of saturating arithmetics converted to normal arithmetics"); STATISTIC(NumNonNull, "Number of function pointer arguments marked non-null"); +STATISTIC(NumCmpIntr, "Number of llvm.[us]cmp intrinsics removed"); STATISTIC(NumMinMax, "Number of llvm.[us]{min,max} intrinsics removed"); STATISTIC(NumSMinMax, "Number of llvm.s{min,max} intrinsics simplified to unsigned"); @@ -548,6 +549,35 @@ static bool processAbsIntrinsic(IntrinsicInst *II, LazyValueInfo *LVI) { return false; } +static bool processCmpIntrinsic(IntrinsicInst *II, LazyValueInfo *LVI) { + bool IsSigned = II->getIntrinsicID() == Intrinsic::scmp; + ConstantRange LHS_CR = LVI->getConstantRangeAtUse(II->getOperandUse(0), + /*UndefAllowed*/ false); + ConstantRange RHS_CR = LVI->getConstantRangeAtUse(II->getOperandUse(1), + /*UndefAllowed*/ false); + + if (LHS_CR.icmp(IsSigned ? ICmpInst::ICMP_SGT : ICmpInst::ICMP_UGT, RHS_CR)) { + ++NumCmpIntr; + II->replaceAllUsesWith(ConstantInt::get(II->getType(), 1)); + II->eraseFromParent(); + return true; + } + if (LHS_CR.icmp(IsSigned ? ICmpInst::ICMP_SLT : ICmpInst::ICMP_ULT, RHS_CR)) { + ++NumCmpIntr; + II->replaceAllUsesWith(ConstantInt::getSigned(II->getType(), -1)); + II->eraseFromParent(); + return true; + } + if (LHS_CR.icmp(ICmpInst::ICMP_EQ, RHS_CR)) { + ++NumCmpIntr; + II->replaceAllUsesWith(ConstantInt::get(II->getType(), 0)); + II->eraseFromParent(); + return true; + } + + return false; +} + // See if this min/max intrinsic always picks it's one specific operand. // If not, check whether we can canonicalize signed minmax into unsigned version static bool processMinMaxIntrinsic(MinMaxIntrinsic *MM, LazyValueInfo *LVI) { @@ -639,6 +669,11 @@ static bool processCallSite(CallBase &CB, LazyValueInfo *LVI) { return processAbsIntrinsic(&cast<IntrinsicInst>(CB), LVI); } + if (CB.getIntrinsicID() == Intrinsic::scmp || + CB.getIntrinsicID() == Intrinsic::ucmp) { + return processCmpIntrinsic(&cast<IntrinsicInst>(CB), LVI); + } + if (auto *MM = dyn_cast<MinMaxIntrinsic>(&CB)) { return processMinMaxIntrinsic(MM, LVI); } diff --git a/llvm/lib/Transforms/Utils/PromoteMemoryToRegister.cpp b/llvm/lib/Transforms/Utils/PromoteMemoryToRegister.cpp index 6f62545..6e021a5 100644 --- a/llvm/lib/Transforms/Utils/PromoteMemoryToRegister.cpp +++ b/llvm/lib/Transforms/Utils/PromoteMemoryToRegister.cpp @@ -41,6 +41,7 @@ #include "llvm/IR/Intrinsics.h" #include "llvm/IR/LLVMContext.h" #include "llvm/IR/Module.h" +#include "llvm/IR/Operator.h" #include "llvm/IR/Type.h" #include "llvm/IR/User.h" #include "llvm/Support/Casting.h" @@ -392,6 +393,9 @@ struct PromoteMem2Reg { /// Lazily compute the number of predecessors a block has. DenseMap<const BasicBlock *, unsigned> BBNumPreds; + /// Whether the function has the no-signed-zeros-fp-math attribute set. + bool NoSignedZeros = false; + public: PromoteMem2Reg(ArrayRef<AllocaInst *> Allocas, DominatorTree &DT, AssumptionCache *AC) @@ -739,6 +743,8 @@ void PromoteMem2Reg::run() { LargeBlockInfo LBI; ForwardIDFCalculator IDF(DT); + NoSignedZeros = F.getFnAttribute("no-signed-zeros-fp-math").getValueAsBool(); + for (unsigned AllocaNum = 0; AllocaNum != Allocas.size(); ++AllocaNum) { AllocaInst *AI = Allocas[AllocaNum]; @@ -1122,6 +1128,14 @@ NextIteration: for (unsigned i = 0; i != NumEdges; ++i) APN->addIncoming(IncomingVals[AllocaNo], Pred); + // For the sequence `return X > 0.0 ? X : -X`, it is expected that this + // results in fabs intrinsic. However, without no-signed-zeros(nsz) flag + // on the phi node generated at this stage, fabs folding does not + // happen. So, we try to infer nsz flag from the function attributes to + // enable this fabs folding. + if (isa<FPMathOperator>(APN) && NoSignedZeros) + APN->setHasNoSignedZeros(true); + // The currently active variable for this block is now the PHI. IncomingVals[AllocaNo] = APN; AllocaATInfo[AllocaNo].updateForNewPhi(APN, DIB); diff --git a/llvm/lib/Transforms/Utils/SimplifyIndVar.cpp b/llvm/lib/Transforms/Utils/SimplifyIndVar.cpp index 74af0ef..e38505c 100644 --- a/llvm/lib/Transforms/Utils/SimplifyIndVar.cpp +++ b/llvm/lib/Transforms/Utils/SimplifyIndVar.cpp @@ -312,6 +312,7 @@ bool SimplifyIndvar::eliminateSDiv(BinaryOperator *SDiv) { SDiv->getName() + ".udiv", SDiv->getIterator()); UDiv->setIsExact(SDiv->isExact()); SDiv->replaceAllUsesWith(UDiv); + UDiv->setDebugLoc(SDiv->getDebugLoc()); LLVM_DEBUG(dbgs() << "INDVARS: Simplified sdiv: " << *SDiv << '\n'); ++NumSimplifiedSDiv; Changed = true; @@ -328,6 +329,7 @@ void SimplifyIndvar::replaceSRemWithURem(BinaryOperator *Rem) { auto *URem = BinaryOperator::Create(BinaryOperator::URem, N, D, Rem->getName() + ".urem", Rem->getIterator()); Rem->replaceAllUsesWith(URem); + URem->setDebugLoc(Rem->getDebugLoc()); LLVM_DEBUG(dbgs() << "INDVARS: Simplified srem: " << *Rem << '\n'); ++NumSimplifiedSRem; Changed = true; @@ -351,6 +353,7 @@ void SimplifyIndvar::replaceRemWithNumeratorOrZero(BinaryOperator *Rem) { SelectInst *Sel = SelectInst::Create(ICmp, ConstantInt::get(T, 0), N, "iv.rem", Rem->getIterator()); Rem->replaceAllUsesWith(Sel); + Sel->setDebugLoc(Rem->getDebugLoc()); LLVM_DEBUG(dbgs() << "INDVARS: Simplified rem: " << *Rem << '\n'); ++NumElimRem; Changed = true; diff --git a/llvm/lib/Transforms/Utils/SplitModule.cpp b/llvm/lib/Transforms/Utils/SplitModule.cpp index 9c39c26d..55db373 100644 --- a/llvm/lib/Transforms/Utils/SplitModule.cpp +++ b/llvm/lib/Transforms/Utils/SplitModule.cpp @@ -105,7 +105,8 @@ static void findPartitions(Module &M, ClusterIDMapType &ClusterIDMap, // At this point module should have the proper mix of globals and locals. // As we attempt to partition this module, we must not change any // locals to globals. - LLVM_DEBUG(dbgs() << "Partition module with (" << M.size() << ")functions\n"); + LLVM_DEBUG(dbgs() << "Partition module with (" << M.size() + << ") functions\n"); ClusterMapType GVtoClusterMap; ComdatMembersType ComdatMembers; @@ -164,10 +165,10 @@ static void findPartitions(Module &M, ClusterIDMapType &ClusterIDMap, std::priority_queue<std::pair<unsigned, unsigned>, std::vector<std::pair<unsigned, unsigned>>, decltype(CompareClusters)> - BalancinQueue(CompareClusters); + BalancingQueue(CompareClusters); // Pre-populate priority queue with N slot blanks. for (unsigned i = 0; i < N; ++i) - BalancinQueue.push(std::make_pair(i, 0)); + BalancingQueue.push(std::make_pair(i, 0)); using SortType = std::pair<unsigned, ClusterMapType::iterator>; @@ -177,11 +178,13 @@ static void findPartitions(Module &M, ClusterIDMapType &ClusterIDMap, // To guarantee determinism, we have to sort SCC according to size. // When size is the same, use leader's name. for (ClusterMapType::iterator I = GVtoClusterMap.begin(), - E = GVtoClusterMap.end(); I != E; ++I) + E = GVtoClusterMap.end(); + I != E; ++I) if (I->isLeader()) Sets.push_back( std::make_pair(std::distance(GVtoClusterMap.member_begin(I), - GVtoClusterMap.member_end()), I)); + GVtoClusterMap.member_end()), + I)); llvm::sort(Sets, [](const SortType &a, const SortType &b) { if (a.first == b.first) @@ -191,9 +194,9 @@ static void findPartitions(Module &M, ClusterIDMapType &ClusterIDMap, }); for (auto &I : Sets) { - unsigned CurrentClusterID = BalancinQueue.top().first; - unsigned CurrentClusterSize = BalancinQueue.top().second; - BalancinQueue.pop(); + unsigned CurrentClusterID = BalancingQueue.top().first; + unsigned CurrentClusterSize = BalancingQueue.top().second; + BalancingQueue.pop(); LLVM_DEBUG(dbgs() << "Root[" << CurrentClusterID << "] cluster_size(" << I.first << ") ----> " << I.second->getData()->getName() @@ -211,7 +214,7 @@ static void findPartitions(Module &M, ClusterIDMapType &ClusterIDMap, CurrentClusterSize++; } // Add this set size to the number of entries in this cluster. - BalancinQueue.push(std::make_pair(CurrentClusterID, CurrentClusterSize)); + BalancingQueue.push(std::make_pair(CurrentClusterID, CurrentClusterSize)); } } @@ -275,8 +278,8 @@ void llvm::SplitModule( ValueToValueMapTy VMap; std::unique_ptr<Module> MPart( CloneModule(M, VMap, [&](const GlobalValue *GV) { - if (ClusterIDMap.count(GV)) - return (ClusterIDMap[GV] == I); + if (auto It = ClusterIDMap.find(GV); It != ClusterIDMap.end()) + return It->second == I; else return isInPartition(GV, I, N); })); diff --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp index c4876bb..868e9e2 100644 --- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp +++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp @@ -835,11 +835,11 @@ static InstructionsState getSameOpcode(ArrayRef<Value *> VL, auto *CallBase = cast<CallInst>(IBase); if (Call->getCalledFunction() != CallBase->getCalledFunction()) return InstructionsState(VL[BaseIndex], nullptr, nullptr); - if (Call->hasOperandBundles() && + if (Call->hasOperandBundles() && (!CallBase->hasOperandBundles() || !std::equal(Call->op_begin() + Call->getBundleOperandsStartIndex(), Call->op_begin() + Call->getBundleOperandsEndIndex(), CallBase->op_begin() + - CallBase->getBundleOperandsStartIndex())) + CallBase->getBundleOperandsStartIndex()))) return InstructionsState(VL[BaseIndex], nullptr, nullptr); Intrinsic::ID ID = getVectorIntrinsicIDForCall(Call, &TLI); if (ID != BaseID) diff --git a/llvm/llvm.spec.in b/llvm/llvm.spec.in deleted file mode 100644 index 8d6ae75..0000000 --- a/llvm/llvm.spec.in +++ /dev/null @@ -1,68 +0,0 @@ -Name: @PACKAGE_NAME@ -Version: @PACKAGE_VERSION@ -Release: 0 -Summary: LLVM (An Optimizing Compiler Infrastructure) -License: Apache-2.0 with LLVM exception -Vendor: None (open source) -Group: Development/Compilers -URL: http://llvm..org/ -Source: http://llvm.org/releases/@PACKAGE_VERSION@/@PACKAGE_NAME@-@PACKAGE_VERSION@.tar.gz -BuildRoot: %{_tmppath}/%{name}-root -Requires: /sbin/ldconfig -BuildRequires: gcc >= 3.4 - -%description -LLVM is a compiler infrastructure designed for compile-time, link-time, runtime, -and idle-time optimization of programs from arbitrary programming languages. -LLVM is written in C++ and has been developed since 2000 at the University of -Illinois and Apple. It currently supports compilation of C and C++ programs, -using front-ends derived from GCC 4.0.1. A new front-end for the C family of -languages is in development. The compiler infrastructure -includes mirror sets of programming tools as well as libraries with equivalent -functionality. - -%prep -%setup -q -n @PACKAGE_NAME@-@PACKAGE_VERSION@ - -%build -./configure \ ---prefix=%{_prefix} \ ---bindir=%{_bindir} \ ---datadir=%{_datadir} \ ---includedir=%{_includedir} \ ---libdir=%{_libdir} \ ---enable-optimized \ ---enable-assertions -make tools-only - -%install -rm -rf %{buildroot} -make install DESTDIR=%{buildroot} - -%clean -rm -rf %{buildroot} - -%post -p /sbin/ldconfig - -%postun -p /sbin/ldconfig - -%files -%defattr(-, root, root) -%doc CREDITS.TXT LICENSE.TXT README.txt docs/*.{html,css,gif,jpg} docs/CommandGuide -%{_bindir}/* -%{_libdir}/*.o -%{_libdir}/*.a -%{_libdir}/*.so -%{_includedir}/llvm - -%changelog -* Fri Aug 04 2006 Reid Spencer -- Updates for release 1.8 -* Fri Apr 07 2006 Reid Spencer -- Make the build be optimized+assertions -* Fri May 13 2005 Reid Spencer -- Minor adjustments for the 1.5 release -* Mon Feb 09 2003 Brian R. Gaeke -- Initial working version of RPM spec file. - - diff --git a/llvm/test/Analysis/ValueTracking/phi-known-bits.ll b/llvm/test/Analysis/ValueTracking/phi-known-bits.ll index 7b5e143..3728e41 100644 --- a/llvm/test/Analysis/ValueTracking/phi-known-bits.ll +++ b/llvm/test/Analysis/ValueTracking/phi-known-bits.ll @@ -375,10 +375,13 @@ F: define i8 @phi_ugt_high_bits_and_known(i8 %xx) { ; CHECK-LABEL: @phi_ugt_high_bits_and_known( ; CHECK-NEXT: entry: -; CHECK-NEXT: [[CMP:%.*]] = icmp ugt i8 [[XX:%.*]], -65 +; CHECK-NEXT: [[X:%.*]] = or i8 [[XX:%.*]], 1 +; CHECK-NEXT: [[CMP:%.*]] = icmp ugt i8 [[XX]], -65 ; CHECK-NEXT: br i1 [[CMP]], label [[T:%.*]], label [[F:%.*]] ; CHECK: T: -; CHECK-NEXT: ret i8 65 +; CHECK-NEXT: [[V:%.*]] = phi i8 [ [[X]], [[ENTRY:%.*]] ], [ -1, [[F]] ] +; CHECK-NEXT: [[R:%.*]] = and i8 [[V]], 65 +; CHECK-NEXT: ret i8 [[R]] ; CHECK: F: ; CHECK-NEXT: br label [[T]] ; diff --git a/llvm/test/CodeGen/AArch64/popcount.ll b/llvm/test/CodeGen/AArch64/popcount.ll index b1231ee..f60d0db 100644 --- a/llvm/test/CodeGen/AArch64/popcount.ll +++ b/llvm/test/CodeGen/AArch64/popcount.ll @@ -1,15 +1,28 @@ ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py -; RUN: llc < %s -O0 -mtriple=aarch64-unknown-unknown | FileCheck %s +; RUN: llc < %s -O0 -mtriple=aarch64-unknown-unknown | FileCheck %s --check-prefix=CHECKO0 +; RUN: llc < %s -mtriple=aarch64-unknown-unknown -mattr=+neon | FileCheck %s --check-prefixes=CHECK,NEON +; RUN: llc < %s -mtriple=aarch64-unknown-unknown -mattr=+neon,+dotprod | FileCheck %s --check-prefixes=CHECK,DOT +; RUN: llc < %s -mtriple=aarch64-unknown-unknown -mattr=+sve | FileCheck %s --check-prefixes=CHECK,SVE ; Function Attrs: nobuiltin nounwind readonly define i8 @popcount128(ptr nocapture nonnull readonly %0) { +; CHECKO0-LABEL: popcount128: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: ldr q0, [x0] +; CHECKO0-NEXT: cnt v0.16b, v0.16b +; CHECKO0-NEXT: uaddlv h0, v0.16b +; CHECKO0-NEXT: // kill: def $q0 killed $h0 +; CHECKO0-NEXT: // kill: def $s0 killed $s0 killed $q0 +; CHECKO0-NEXT: fmov w0, s0 +; CHECKO0-NEXT: ret +; ; CHECK-LABEL: popcount128: ; CHECK: // %bb.0: // %Entry -; CHECK-NEXT: ldr q0, [x0] +; CHECK-NEXT: ldr d0, [x0] +; CHECK-NEXT: add x8, x0, #8 +; CHECK-NEXT: ld1 { v0.d }[1], [x8] ; CHECK-NEXT: cnt v0.16b, v0.16b ; CHECK-NEXT: uaddlv h0, v0.16b -; CHECK-NEXT: // kill: def $q0 killed $h0 -; CHECK-NEXT: // kill: def $s0 killed $s0 killed $q0 ; CHECK-NEXT: fmov w0, s0 ; CHECK-NEXT: ret Entry: @@ -24,37 +37,54 @@ declare i128 @llvm.ctpop.i128(i128) ; Function Attrs: nobuiltin nounwind readonly define i16 @popcount256(ptr nocapture nonnull readonly %0) { +; CHECKO0-LABEL: popcount256: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: ldr x11, [x0] +; CHECKO0-NEXT: ldr x10, [x0, #8] +; CHECKO0-NEXT: ldr x9, [x0, #16] +; CHECKO0-NEXT: ldr x8, [x0, #24] +; CHECKO0-NEXT: // implicit-def: $q1 +; CHECKO0-NEXT: mov v1.d[0], x11 +; CHECKO0-NEXT: mov v1.d[1], x10 +; CHECKO0-NEXT: // implicit-def: $q0 +; CHECKO0-NEXT: mov v0.d[0], x9 +; CHECKO0-NEXT: mov v0.d[1], x8 +; CHECKO0-NEXT: cnt v1.16b, v1.16b +; CHECKO0-NEXT: uaddlv h1, v1.16b +; CHECKO0-NEXT: // kill: def $q1 killed $h1 +; CHECKO0-NEXT: // kill: def $s1 killed $s1 killed $q1 +; CHECKO0-NEXT: fmov w0, s1 +; CHECKO0-NEXT: mov w10, wzr +; CHECKO0-NEXT: mov w9, w0 +; CHECKO0-NEXT: mov w8, w10 +; CHECKO0-NEXT: bfi x9, x8, #32, #32 +; CHECKO0-NEXT: cnt v0.16b, v0.16b +; CHECKO0-NEXT: uaddlv h0, v0.16b +; CHECKO0-NEXT: // kill: def $q0 killed $h0 +; CHECKO0-NEXT: // kill: def $s0 killed $s0 killed $q0 +; CHECKO0-NEXT: fmov w0, s0 +; CHECKO0-NEXT: mov w8, w0 +; CHECKO0-NEXT: // kill: def $x10 killed $w10 +; CHECKO0-NEXT: bfi x8, x10, #32, #32 +; CHECKO0-NEXT: adds x8, x8, x9 +; CHECKO0-NEXT: mov w0, w8 +; CHECKO0-NEXT: ret +; ; CHECK-LABEL: popcount256: ; CHECK: // %bb.0: // %Entry -; CHECK-NEXT: ldr x11, [x0] -; CHECK-NEXT: ldr x10, [x0, #8] -; CHECK-NEXT: ldr x9, [x0, #16] -; CHECK-NEXT: ldr x8, [x0, #24] -; CHECK-NEXT: // implicit-def: $q1 -; CHECK-NEXT: mov v1.d[0], x11 -; CHECK-NEXT: mov v1.d[1], x10 -; CHECK-NEXT: // implicit-def: $q0 -; CHECK-NEXT: mov v0.d[0], x9 -; CHECK-NEXT: mov v0.d[1], x8 -; CHECK-NEXT: cnt v1.16b, v1.16b -; CHECK-NEXT: uaddlv h1, v1.16b -; CHECK-NEXT: // kill: def $q1 killed $h1 -; CHECK-NEXT: // kill: def $s1 killed $s1 killed $q1 -; CHECK-NEXT: fmov w0, s1 -; CHECK-NEXT: mov w10, wzr -; CHECK-NEXT: mov w9, w0 -; CHECK-NEXT: mov w8, w10 -; CHECK-NEXT: bfi x9, x8, #32, #32 +; CHECK-NEXT: ldr d0, [x0, #16] +; CHECK-NEXT: ldr d1, [x0] +; CHECK-NEXT: add x8, x0, #8 +; CHECK-NEXT: add x9, x0, #24 +; CHECK-NEXT: ld1 { v0.d }[1], [x9] +; CHECK-NEXT: ld1 { v1.d }[1], [x8] ; CHECK-NEXT: cnt v0.16b, v0.16b +; CHECK-NEXT: cnt v1.16b, v1.16b ; CHECK-NEXT: uaddlv h0, v0.16b -; CHECK-NEXT: // kill: def $q0 killed $h0 -; CHECK-NEXT: // kill: def $s0 killed $s0 killed $q0 -; CHECK-NEXT: fmov w0, s0 -; CHECK-NEXT: mov w8, w0 -; CHECK-NEXT: // kill: def $x10 killed $w10 -; CHECK-NEXT: bfi x8, x10, #32, #32 -; CHECK-NEXT: adds x8, x8, x9 -; CHECK-NEXT: mov w0, w8 +; CHECK-NEXT: uaddlv h1, v1.16b +; CHECK-NEXT: fmov w8, s0 +; CHECK-NEXT: fmov w9, s1 +; CHECK-NEXT: add w0, w9, w8 ; CHECK-NEXT: ret Entry: %1 = load i256, ptr %0, align 16 @@ -67,25 +97,215 @@ Entry: declare i256 @llvm.ctpop.i256(i256) define <1 x i128> @popcount1x128(<1 x i128> %0) { +; CHECKO0-LABEL: popcount1x128: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: // implicit-def: $q0 +; CHECKO0-NEXT: mov v0.d[0], x0 +; CHECKO0-NEXT: mov v0.d[1], x1 +; CHECKO0-NEXT: cnt v0.16b, v0.16b +; CHECKO0-NEXT: uaddlv h0, v0.16b +; CHECKO0-NEXT: // kill: def $q0 killed $h0 +; CHECKO0-NEXT: mov x1, xzr +; CHECKO0-NEXT: // kill: def $s0 killed $s0 killed $q0 +; CHECKO0-NEXT: fmov w0, s0 +; CHECKO0-NEXT: mov w8, wzr +; CHECKO0-NEXT: // kill: def $x0 killed $w0 +; CHECKO0-NEXT: // kill: def $x8 killed $w8 +; CHECKO0-NEXT: bfi x0, x8, #32, #32 +; CHECKO0-NEXT: ret +; ; CHECK-LABEL: popcount1x128: ; CHECK: // %bb.0: // %Entry -; CHECK-NEXT: // implicit-def: $q0 -; CHECK-NEXT: mov v0.d[0], x0 -; CHECK-NEXT: mov v0.d[1], x1 -; CHECK-NEXT: cnt v0.16b, v0.16b -; CHECK-NEXT: uaddlv h0, v0.16b -; CHECK-NEXT: // kill: def $q0 killed $h0 -; CHECK-NEXT: mov x1, xzr -; CHECK-NEXT: // kill: def $s0 killed $s0 killed $q0 -; CHECK-NEXT: fmov w0, s0 -; CHECK-NEXT: mov w8, wzr -; CHECK-NEXT: // kill: def $x0 killed $w0 -; CHECK-NEXT: // kill: def $x8 killed $w8 -; CHECK-NEXT: bfi x0, x8, #32, #32 +; CHECK-NEXT: fmov d1, x0 +; CHECK-NEXT: movi v0.2d, #0000000000000000 +; CHECK-NEXT: mov v1.d[1], x1 +; CHECK-NEXT: cnt v1.16b, v1.16b +; CHECK-NEXT: uaddlv h1, v1.16b +; CHECK-NEXT: mov v0.s[0], v1.s[0] +; CHECK-NEXT: mov x1, v0.d[1] +; CHECK-NEXT: fmov x0, d0 ; CHECK-NEXT: ret Entry: - %1 = tail call <1 x i128> @llvm.ctpop.v1.i128(<1 x i128> %0) + %1 = tail call <1 x i128> @llvm.ctpop.v1i128(<1 x i128> %0) ret <1 x i128> %1 } -declare <1 x i128> @llvm.ctpop.v1.i128(<1 x i128>) +declare <1 x i128> @llvm.ctpop.v1i128(<1 x i128>) + +define <2 x i64> @popcount2x64(<2 x i64> %0) { +; CHECKO0-LABEL: popcount2x64: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: cnt v0.16b, v0.16b +; CHECKO0-NEXT: uaddlp v0.8h, v0.16b +; CHECKO0-NEXT: uaddlp v0.4s, v0.8h +; CHECKO0-NEXT: uaddlp v0.2d, v0.4s +; CHECKO0-NEXT: ret +; +; NEON-LABEL: popcount2x64: +; NEON: // %bb.0: // %Entry +; NEON-NEXT: cnt v0.16b, v0.16b +; NEON-NEXT: uaddlp v0.8h, v0.16b +; NEON-NEXT: uaddlp v0.4s, v0.8h +; NEON-NEXT: uaddlp v0.2d, v0.4s +; NEON-NEXT: ret +; +; DOT-LABEL: popcount2x64: +; DOT: // %bb.0: // %Entry +; DOT-NEXT: movi v1.16b, #1 +; DOT-NEXT: cnt v0.16b, v0.16b +; DOT-NEXT: movi v2.2d, #0000000000000000 +; DOT-NEXT: udot v2.4s, v1.16b, v0.16b +; DOT-NEXT: uaddlp v0.2d, v2.4s +; DOT-NEXT: ret +; +; SVE-LABEL: popcount2x64: +; SVE: // %bb.0: // %Entry +; SVE-NEXT: cnt v0.16b, v0.16b +; SVE-NEXT: uaddlp v0.8h, v0.16b +; SVE-NEXT: uaddlp v0.4s, v0.8h +; SVE-NEXT: uaddlp v0.2d, v0.4s +; SVE-NEXT: ret +Entry: + %1 = tail call <2 x i64> @llvm.ctpop.v2i64(<2 x i64> %0) + ret <2 x i64> %1 +} + +declare <2 x i64> @llvm.ctpop.v2i64(<2 x i64>) + +define <1 x i64> @popcount1x64(<1 x i64> %0) { +; CHECKO0-LABEL: popcount1x64: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: fmov x0, d0 +; CHECKO0-NEXT: fmov d0, x0 +; CHECKO0-NEXT: cnt v0.8b, v0.8b +; CHECKO0-NEXT: uaddlv h0, v0.8b +; CHECKO0-NEXT: // kill: def $q0 killed $h0 +; CHECKO0-NEXT: mov w8, v0.s[0] +; CHECKO0-NEXT: // kill: def $x8 killed $w8 +; CHECKO0-NEXT: fmov d0, x8 +; CHECKO0-NEXT: ret +; +; CHECK-LABEL: popcount1x64: +; CHECK: // %bb.0: // %Entry +; CHECK-NEXT: cnt v0.8b, v0.8b +; CHECK-NEXT: uaddlp v0.4h, v0.8b +; CHECK-NEXT: uaddlp v0.2s, v0.4h +; CHECK-NEXT: uaddlp v0.1d, v0.2s +; CHECK-NEXT: ret +Entry: + %1 = tail call <1 x i64> @llvm.ctpop.v1i64(<1 x i64> %0) + ret <1 x i64> %1 +} + +declare <1 x i64> @llvm.ctpop.v1i64(<1 x i64>) + +define <4 x i32> @popcount4x32(<4 x i32> %0) { +; CHECKO0-LABEL: popcount4x32: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: cnt v0.16b, v0.16b +; CHECKO0-NEXT: uaddlp v0.8h, v0.16b +; CHECKO0-NEXT: uaddlp v0.4s, v0.8h +; CHECKO0-NEXT: ret +; +; NEON-LABEL: popcount4x32: +; NEON: // %bb.0: // %Entry +; NEON-NEXT: cnt v0.16b, v0.16b +; NEON-NEXT: uaddlp v0.8h, v0.16b +; NEON-NEXT: uaddlp v0.4s, v0.8h +; NEON-NEXT: ret +; +; DOT-LABEL: popcount4x32: +; DOT: // %bb.0: // %Entry +; DOT-NEXT: movi v1.16b, #1 +; DOT-NEXT: cnt v2.16b, v0.16b +; DOT-NEXT: movi v0.2d, #0000000000000000 +; DOT-NEXT: udot v0.4s, v1.16b, v2.16b +; DOT-NEXT: ret +; +; SVE-LABEL: popcount4x32: +; SVE: // %bb.0: // %Entry +; SVE-NEXT: cnt v0.16b, v0.16b +; SVE-NEXT: uaddlp v0.8h, v0.16b +; SVE-NEXT: uaddlp v0.4s, v0.8h +; SVE-NEXT: ret +Entry: + %1 = tail call <4 x i32> @llvm.ctpop.v4i32(<4 x i32> %0) + ret <4 x i32> %1 +} + +declare <4 x i32> @llvm.ctpop.v4i32(<4 x i32>) + +define <2 x i32> @popcount2x32(<2 x i32> %0) { +; CHECKO0-LABEL: popcount2x32: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: cnt v0.8b, v0.8b +; CHECKO0-NEXT: uaddlp v0.4h, v0.8b +; CHECKO0-NEXT: uaddlp v0.2s, v0.4h +; CHECKO0-NEXT: ret +; +; NEON-LABEL: popcount2x32: +; NEON: // %bb.0: // %Entry +; NEON-NEXT: cnt v0.8b, v0.8b +; NEON-NEXT: uaddlp v0.4h, v0.8b +; NEON-NEXT: uaddlp v0.2s, v0.4h +; NEON-NEXT: ret +; +; DOT-LABEL: popcount2x32: +; DOT: // %bb.0: // %Entry +; DOT-NEXT: movi v1.2d, #0000000000000000 +; DOT-NEXT: cnt v0.8b, v0.8b +; DOT-NEXT: movi v2.8b, #1 +; DOT-NEXT: udot v1.2s, v2.8b, v0.8b +; DOT-NEXT: fmov d0, d1 +; DOT-NEXT: ret +; +; SVE-LABEL: popcount2x32: +; SVE: // %bb.0: // %Entry +; SVE-NEXT: cnt v0.8b, v0.8b +; SVE-NEXT: uaddlp v0.4h, v0.8b +; SVE-NEXT: uaddlp v0.2s, v0.4h +; SVE-NEXT: ret +Entry: + %1 = tail call <2 x i32> @llvm.ctpop.v2i32(<2 x i32> %0) + ret <2 x i32> %1 +} + +declare <2 x i32> @llvm.ctpop.v2i32(<2 x i32>) + +define <8 x i16> @popcount8x16(<8 x i16> %0) { +; CHECKO0-LABEL: popcount8x16: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: cnt v0.16b, v0.16b +; CHECKO0-NEXT: uaddlp v0.8h, v0.16b +; CHECKO0-NEXT: ret +; +; CHECK-LABEL: popcount8x16: +; CHECK: // %bb.0: // %Entry +; CHECK-NEXT: cnt v0.16b, v0.16b +; CHECK-NEXT: uaddlp v0.8h, v0.16b +; CHECK-NEXT: ret +Entry: + %1 = tail call <8 x i16> @llvm.ctpop.v8i16(<8 x i16> %0) + ret <8 x i16> %1 +} + +declare <8 x i16> @llvm.ctpop.v8i16(<8 x i16>) + +define <4 x i16> @popcount4x16(<4 x i16> %0) { +; CHECKO0-LABEL: popcount4x16: +; CHECKO0: // %bb.0: // %Entry +; CHECKO0-NEXT: cnt v0.8b, v0.8b +; CHECKO0-NEXT: uaddlp v0.4h, v0.8b +; CHECKO0-NEXT: ret +; +; CHECK-LABEL: popcount4x16: +; CHECK: // %bb.0: // %Entry +; CHECK-NEXT: cnt v0.8b, v0.8b +; CHECK-NEXT: uaddlp v0.4h, v0.8b +; CHECK-NEXT: ret +Entry: + %1 = tail call <4 x i16> @llvm.ctpop.v4i16(<4 x i16> %0) + ret <4 x i16> %1 +} + +declare <4 x i16> @llvm.ctpop.v4i16(<4 x i16>) diff --git a/llvm/test/CodeGen/AMDGPU/global_atomic_optimizer_fp_rtn.ll b/llvm/test/CodeGen/AMDGPU/global_atomic_optimizer_fp_rtn.ll index 722c069..fca86a91 100644 --- a/llvm/test/CodeGen/AMDGPU/global_atomic_optimizer_fp_rtn.ll +++ b/llvm/test/CodeGen/AMDGPU/global_atomic_optimizer_fp_rtn.ll @@ -506,7 +506,7 @@ define amdgpu_ps float @global_atomic_fmin_uni_address_uni_value_agent_scope_uns ; IR-NEXT: [[TMP15:%.*]] = call i32 @llvm.amdgcn.readfirstlane.i32(i32 [[TMP14]]) ; IR-NEXT: [[TMP16:%.*]] = bitcast i32 [[TMP15]] to float ; IR-NEXT: [[TMP17:%.*]] = uitofp i32 [[TMP8]] to float -; IR-NEXT: [[TMP18:%.*]] = select i1 [[TMP9]], float 0x7FF0000000000000, float [[VAL]] +; IR-NEXT: [[TMP18:%.*]] = select i1 [[TMP9]], float 0x7FF8000000000000, float [[VAL]] ; IR-NEXT: [[TMP19:%.*]] = call float @llvm.minnum.f32(float [[TMP16]], float [[TMP18]]) ; IR-NEXT: br label [[TMP20]] ; IR: 20: @@ -544,7 +544,7 @@ define amdgpu_ps float @global_atomic_fmin_uni_address_div_value_agent_scope_uns ; IR-ITERATIVE-NEXT: [[TMP19:%.*]] = phi float [ poison, [[TMP0:%.*]] ], [ [[TMP17]], [[TMP12]] ] ; IR-ITERATIVE-NEXT: ret float [[TMP19]] ; IR-ITERATIVE: ComputeLoop: -; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0x7FF0000000000000, [[TMP2]] ], [ [[TMP29]], [[COMPUTELOOP]] ] +; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0x7FF8000000000000, [[TMP2]] ], [ [[TMP29]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[OLDVALUEPHI:%.*]] = phi float [ poison, [[TMP2]] ], [ [[TMP28]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[ACTIVEBITS:%.*]] = phi i64 [ [[TMP9]], [[TMP2]] ], [ [[TMP32:%.*]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[TMP20:%.*]] = call i64 @llvm.cttz.i64(i64 [[ACTIVEBITS]], i1 true) @@ -577,22 +577,22 @@ define amdgpu_ps float @global_atomic_fmin_uni_address_div_value_agent_scope_uns ; IR-DPP-NEXT: [[TMP7:%.*]] = call i32 @llvm.amdgcn.mbcnt.lo(i32 [[TMP4]], i32 0) ; IR-DPP-NEXT: [[TMP8:%.*]] = call i32 @llvm.amdgcn.mbcnt.hi(i32 [[TMP6]], i32 [[TMP7]]) ; IR-DPP-NEXT: [[TMP9:%.*]] = bitcast float [[VAL:%.*]] to i32 -; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 2139095040) +; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 2143289344) ; IR-DPP-NEXT: [[TMP11:%.*]] = bitcast i32 [[TMP10]] to float ; IR-DPP-NEXT: [[TMP12:%.*]] = bitcast i32 [[TMP9]] to float -; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP14:%.*]] = call float @llvm.minnum.f32(float [[TMP11]], float [[TMP13]]) -; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP16:%.*]] = call float @llvm.minnum.f32(float [[TMP14]], float [[TMP15]]) -; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP18:%.*]] = call float @llvm.minnum.f32(float [[TMP16]], float [[TMP17]]) -; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP20:%.*]] = call float @llvm.minnum.f32(float [[TMP18]], float [[TMP19]]) -; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP22:%.*]] = call float @llvm.minnum.f32(float [[TMP20]], float [[TMP21]]) -; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP24:%.*]] = call float @llvm.minnum.f32(float [[TMP22]], float [[TMP23]]) -; IR-DPP-NEXT: [[TMP25:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP24]], i32 312, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP25:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP24]], i32 312, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP26:%.*]] = bitcast float [[TMP24]] to i32 ; IR-DPP-NEXT: [[TMP27:%.*]] = call i32 @llvm.amdgcn.readlane.i32(i32 [[TMP26]], i32 63) ; IR-DPP-NEXT: [[TMP28:%.*]] = bitcast i32 [[TMP27]] to float @@ -640,7 +640,7 @@ define amdgpu_ps float @global_atomic_fmax_uni_address_uni_value_agent_scope_uns ; IR-ITERATIVE-NEXT: [[TMP15:%.*]] = call i32 @llvm.amdgcn.readfirstlane.i32(i32 [[TMP14]]) #[[ATTR7]] ; IR-ITERATIVE-NEXT: [[TMP16:%.*]] = bitcast i32 [[TMP15]] to float ; IR-ITERATIVE-NEXT: [[TMP17:%.*]] = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 [[TMP8]], metadata !"round.dynamic", metadata !"fpexcept.strict") #[[ATTR7]] -; IR-ITERATIVE-NEXT: [[TMP18:%.*]] = select i1 [[TMP9]], float 0xFFF0000000000000, float [[VAL]] +; IR-ITERATIVE-NEXT: [[TMP18:%.*]] = select i1 [[TMP9]], float 0x7FF8000000000000, float [[VAL]] ; IR-ITERATIVE-NEXT: [[TMP19:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP16]], float [[TMP18]], metadata !"fpexcept.strict") #[[ATTR7]] ; IR-ITERATIVE-NEXT: br label [[TMP20]] ; IR-ITERATIVE: 20: @@ -668,7 +668,7 @@ define amdgpu_ps float @global_atomic_fmax_uni_address_uni_value_agent_scope_uns ; IR-DPP-NEXT: [[TMP15:%.*]] = call i32 @llvm.amdgcn.readfirstlane.i32(i32 [[TMP14]]) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP16:%.*]] = bitcast i32 [[TMP15]] to float ; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.experimental.constrained.uitofp.f32.i32(i32 [[TMP8]], metadata !"round.dynamic", metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP18:%.*]] = select i1 [[TMP9]], float 0xFFF0000000000000, float [[VAL]] +; IR-DPP-NEXT: [[TMP18:%.*]] = select i1 [[TMP9]], float 0x7FF8000000000000, float [[VAL]] ; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP16]], float [[TMP18]], metadata !"fpexcept.strict") #[[ATTR8]] ; IR-DPP-NEXT: br label [[TMP20]] ; IR-DPP: 20: @@ -706,7 +706,7 @@ define amdgpu_ps float @global_atomic_fmax_uni_address_div_value_agent_scope_uns ; IR-ITERATIVE-NEXT: [[TMP19:%.*]] = phi float [ poison, [[TMP0:%.*]] ], [ [[TMP17]], [[TMP12]] ] ; IR-ITERATIVE-NEXT: ret float [[TMP19]] ; IR-ITERATIVE: ComputeLoop: -; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0xFFF0000000000000, [[TMP2]] ], [ [[TMP29]], [[COMPUTELOOP]] ] +; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0x7FF8000000000000, [[TMP2]] ], [ [[TMP29]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[OLDVALUEPHI:%.*]] = phi float [ poison, [[TMP2]] ], [ [[TMP28]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[ACTIVEBITS:%.*]] = phi i64 [ [[TMP9]], [[TMP2]] ], [ [[TMP32:%.*]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[TMP20:%.*]] = call i64 @llvm.cttz.i64(i64 [[ACTIVEBITS]], i1 true) #[[ATTR7]] @@ -739,22 +739,22 @@ define amdgpu_ps float @global_atomic_fmax_uni_address_div_value_agent_scope_uns ; IR-DPP-NEXT: [[TMP7:%.*]] = call i32 @llvm.amdgcn.mbcnt.lo(i32 [[TMP4]], i32 0) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP8:%.*]] = call i32 @llvm.amdgcn.mbcnt.hi(i32 [[TMP6]], i32 [[TMP7]]) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP9:%.*]] = bitcast float [[VAL:%.*]] to i32 -; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 -8388608) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 2143289344) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP11:%.*]] = bitcast i32 [[TMP10]] to float ; IR-DPP-NEXT: [[TMP12:%.*]] = bitcast i32 [[TMP9]] to float -; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP14:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP11]], float [[TMP13]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP16:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP14]], float [[TMP15]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP18:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP16]], float [[TMP17]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP20:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP18]], float [[TMP19]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP22:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP20]], float [[TMP21]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP24:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP22]], float [[TMP23]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP25:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP24]], i32 312, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP25:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP24]], i32 312, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP26:%.*]] = bitcast float [[TMP24]] to i32 ; IR-DPP-NEXT: [[TMP27:%.*]] = call i32 @llvm.amdgcn.readlane.i32(i32 [[TMP26]], i32 63) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP28:%.*]] = bitcast i32 [[TMP27]] to float @@ -1318,7 +1318,7 @@ define amdgpu_ps double @global_atomic_fmin_double_uni_address_uni_value_agent_s ; IR-NEXT: [[TMP21:%.*]] = insertelement <2 x i32> [[TMP20]], i32 [[TMP19]], i32 1 ; IR-NEXT: [[TMP22:%.*]] = bitcast <2 x i32> [[TMP21]] to double ; IR-NEXT: [[TMP23:%.*]] = uitofp i32 [[TMP8]] to double -; IR-NEXT: [[TMP24:%.*]] = select i1 [[TMP9]], double 0x7FF0000000000000, double [[VAL]] +; IR-NEXT: [[TMP24:%.*]] = select i1 [[TMP9]], double 0x7FF8000000000000, double [[VAL]] ; IR-NEXT: [[TMP25:%.*]] = call double @llvm.minnum.f64(double [[TMP22]], double [[TMP24]]) ; IR-NEXT: br label [[TMP26]] ; IR: 26: @@ -1366,7 +1366,7 @@ define amdgpu_ps double @global_atomic__fmax_double_uni_address_uni_value_agent_ ; IR-ITERATIVE-NEXT: [[TMP21:%.*]] = insertelement <2 x i32> [[TMP20]], i32 [[TMP19]], i32 1 ; IR-ITERATIVE-NEXT: [[TMP22:%.*]] = bitcast <2 x i32> [[TMP21]] to double ; IR-ITERATIVE-NEXT: [[TMP23:%.*]] = call double @llvm.experimental.constrained.uitofp.f64.i32(i32 [[TMP8]], metadata !"round.dynamic", metadata !"fpexcept.strict") #[[ATTR7]] -; IR-ITERATIVE-NEXT: [[TMP24:%.*]] = select i1 [[TMP9]], double 0xFFF0000000000000, double [[VAL]] +; IR-ITERATIVE-NEXT: [[TMP24:%.*]] = select i1 [[TMP9]], double 0x7FF8000000000000, double [[VAL]] ; IR-ITERATIVE-NEXT: [[TMP25:%.*]] = call double @llvm.experimental.constrained.maxnum.f64(double [[TMP22]], double [[TMP24]], metadata !"fpexcept.strict") #[[ATTR7]] ; IR-ITERATIVE-NEXT: br label [[TMP26]] ; IR-ITERATIVE: 26: @@ -1400,7 +1400,7 @@ define amdgpu_ps double @global_atomic__fmax_double_uni_address_uni_value_agent_ ; IR-DPP-NEXT: [[TMP21:%.*]] = insertelement <2 x i32> [[TMP20]], i32 [[TMP19]], i32 1 ; IR-DPP-NEXT: [[TMP22:%.*]] = bitcast <2 x i32> [[TMP21]] to double ; IR-DPP-NEXT: [[TMP23:%.*]] = call double @llvm.experimental.constrained.uitofp.f64.i32(i32 [[TMP8]], metadata !"round.dynamic", metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP24:%.*]] = select i1 [[TMP9]], double 0xFFF0000000000000, double [[VAL]] +; IR-DPP-NEXT: [[TMP24:%.*]] = select i1 [[TMP9]], double 0x7FF8000000000000, double [[VAL]] ; IR-DPP-NEXT: [[TMP25:%.*]] = call double @llvm.experimental.constrained.maxnum.f64(double [[TMP22]], double [[TMP24]], metadata !"fpexcept.strict") #[[ATTR8]] ; IR-DPP-NEXT: br label [[TMP26]] ; IR-DPP: 26: diff --git a/llvm/test/CodeGen/AMDGPU/global_atomics_optimizer_fp_no_rtn.ll b/llvm/test/CodeGen/AMDGPU/global_atomics_optimizer_fp_no_rtn.ll index fa66a0f..3cf2a85b 100644 --- a/llvm/test/CodeGen/AMDGPU/global_atomics_optimizer_fp_no_rtn.ll +++ b/llvm/test/CodeGen/AMDGPU/global_atomics_optimizer_fp_no_rtn.ll @@ -433,7 +433,7 @@ define amdgpu_ps void @global_atomic_fmin_uni_address_div_value_agent_scope_unsa ; IR-ITERATIVE: 13: ; IR-ITERATIVE-NEXT: ret void ; IR-ITERATIVE: ComputeLoop: -; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0x7FF0000000000000, [[TMP2]] ], [ [[TMP19]], [[COMPUTELOOP]] ] +; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0x7FF8000000000000, [[TMP2]] ], [ [[TMP19]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[ACTIVEBITS:%.*]] = phi i64 [ [[TMP9]], [[TMP2]] ], [ [[TMP22:%.*]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[TMP14:%.*]] = call i64 @llvm.cttz.i64(i64 [[ACTIVEBITS]], i1 true) ; IR-ITERATIVE-NEXT: [[TMP15:%.*]] = trunc i64 [[TMP14]] to i32 @@ -461,20 +461,20 @@ define amdgpu_ps void @global_atomic_fmin_uni_address_div_value_agent_scope_unsa ; IR-DPP-NEXT: [[TMP7:%.*]] = call i32 @llvm.amdgcn.mbcnt.lo(i32 [[TMP4]], i32 0) ; IR-DPP-NEXT: [[TMP8:%.*]] = call i32 @llvm.amdgcn.mbcnt.hi(i32 [[TMP6]], i32 [[TMP7]]) ; IR-DPP-NEXT: [[TMP9:%.*]] = bitcast float [[VAL:%.*]] to i32 -; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 2139095040) +; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 2143289344) ; IR-DPP-NEXT: [[TMP11:%.*]] = bitcast i32 [[TMP10]] to float ; IR-DPP-NEXT: [[TMP12:%.*]] = bitcast i32 [[TMP9]] to float -; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP14:%.*]] = call float @llvm.minnum.f32(float [[TMP11]], float [[TMP13]]) -; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP16:%.*]] = call float @llvm.minnum.f32(float [[TMP14]], float [[TMP15]]) -; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP18:%.*]] = call float @llvm.minnum.f32(float [[TMP16]], float [[TMP17]]) -; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP20:%.*]] = call float @llvm.minnum.f32(float [[TMP18]], float [[TMP19]]) -; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP22:%.*]] = call float @llvm.minnum.f32(float [[TMP20]], float [[TMP21]]) -; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF0000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) +; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) ; IR-DPP-NEXT: [[TMP24:%.*]] = call float @llvm.minnum.f32(float [[TMP22]], float [[TMP23]]) ; IR-DPP-NEXT: [[TMP25:%.*]] = bitcast float [[TMP24]] to i32 ; IR-DPP-NEXT: [[TMP26:%.*]] = call i32 @llvm.amdgcn.readlane.i32(i32 [[TMP25]], i32 63) @@ -560,7 +560,7 @@ define amdgpu_ps void @global_atomic_fmax_uni_address_div_value_agent_scope_unsa ; IR-ITERATIVE: 13: ; IR-ITERATIVE-NEXT: ret void ; IR-ITERATIVE: ComputeLoop: -; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0xFFF0000000000000, [[TMP2]] ], [ [[TMP19]], [[COMPUTELOOP]] ] +; IR-ITERATIVE-NEXT: [[ACCUMULATOR:%.*]] = phi float [ 0x7FF8000000000000, [[TMP2]] ], [ [[TMP19]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[ACTIVEBITS:%.*]] = phi i64 [ [[TMP9]], [[TMP2]] ], [ [[TMP22:%.*]], [[COMPUTELOOP]] ] ; IR-ITERATIVE-NEXT: [[TMP14:%.*]] = call i64 @llvm.cttz.i64(i64 [[ACTIVEBITS]], i1 true) #[[ATTR7]] ; IR-ITERATIVE-NEXT: [[TMP15:%.*]] = trunc i64 [[TMP14]] to i32 @@ -588,20 +588,20 @@ define amdgpu_ps void @global_atomic_fmax_uni_address_div_value_agent_scope_unsa ; IR-DPP-NEXT: [[TMP7:%.*]] = call i32 @llvm.amdgcn.mbcnt.lo(i32 [[TMP4]], i32 0) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP8:%.*]] = call i32 @llvm.amdgcn.mbcnt.hi(i32 [[TMP6]], i32 [[TMP7]]) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP9:%.*]] = bitcast float [[VAL:%.*]] to i32 -; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 -8388608) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.set.inactive.i32(i32 [[TMP9]], i32 2143289344) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP11:%.*]] = bitcast i32 [[TMP10]] to float ; IR-DPP-NEXT: [[TMP12:%.*]] = bitcast i32 [[TMP9]] to float -; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP13:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP11]], i32 273, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP14:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP11]], float [[TMP13]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP15:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP14]], i32 274, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP16:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP14]], float [[TMP15]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP17:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP16]], i32 276, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP18:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP16]], float [[TMP17]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP19:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP18]], i32 280, i32 15, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP20:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP18]], float [[TMP19]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP21:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP20]], i32 322, i32 10, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP22:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP20]], float [[TMP21]], metadata !"fpexcept.strict") #[[ATTR8]] -; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0xFFF0000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) #[[ATTR8]] +; IR-DPP-NEXT: [[TMP23:%.*]] = call float @llvm.amdgcn.update.dpp.f32(float 0x7FF8000000000000, float [[TMP22]], i32 323, i32 12, i32 15, i1 false) #[[ATTR8]] ; IR-DPP-NEXT: [[TMP24:%.*]] = call float @llvm.experimental.constrained.maxnum.f32(float [[TMP22]], float [[TMP23]], metadata !"fpexcept.strict") #[[ATTR8]] ; IR-DPP-NEXT: [[TMP25:%.*]] = bitcast float [[TMP24]] to i32 ; IR-DPP-NEXT: [[TMP26:%.*]] = call i32 @llvm.amdgcn.readlane.i32(i32 [[TMP25]], i32 63) #[[ATTR8]] diff --git a/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmax.ll b/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmax.ll index 7f052e1..e02d749 100644 --- a/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmax.ll +++ b/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmax.ll @@ -326,7 +326,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX9-NEXT: s_waitcnt lgkmcnt(0) ; GFX9-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX9-NEXT: s_mov_b64 s[0:1], exec -; GFX9-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX9-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX9-NEXT: .LBB1_1: ; %ComputeLoop ; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1 ; GFX9-NEXT: s_ff1_i32_b64 s2, s[0:1] @@ -395,7 +395,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1064-NEXT: s_mov_b32 s32, 0 ; GFX1064-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1064-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1064-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1064-NEXT: s_mov_b64 s[0:1], exec ; GFX1064-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1064-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -451,7 +451,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1032-NEXT: s_mov_b32 s32, 0 ; GFX1032-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1032-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1032-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1032-NEXT: s_mov_b32 s0, exec_lo ; GFX1032-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1032-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -496,7 +496,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1164-NEXT: s_mov_b32 s32, 0 ; GFX1164-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1164-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1164-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-NEXT: s_mov_b64 s[0:1], exec ; GFX1164-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1164-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -546,7 +546,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1132-NEXT: s_mov_b32 s32, 0 ; GFX1132-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1132-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1132-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-NEXT: s_mov_b32 s0, exec_lo ; GFX1132-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1132-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -609,32 +609,32 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX9-DPP-NEXT: v_mbcnt_hi_u32_b32 v1, exec_hi, v1 ; GFX9-DPP-NEXT: v_mov_b32_e32 v3, v0 ; GFX9-DPP-NEXT: s_not_b64 exec, exec -; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX9-DPP-NEXT: s_not_b64 exec, exec ; GFX9-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 -; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 0 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:1 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:2 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:4 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:8 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_bcast:15 row_mask:0xa bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 @@ -699,25 +699,25 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1064-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1064-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec -; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v4, v3 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -777,25 +777,25 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1032-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1032-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v4, v3 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -837,30 +837,30 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1164-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec -; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1164-DPP-NEXT: s_waitcnt_depctr 0xfff ; GFX1164-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v2, v2 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v2, v1 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -916,26 +916,26 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_agent_scope_ ; GFX1132-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v2, v1 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -1386,7 +1386,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX9-NEXT: s_waitcnt lgkmcnt(0) ; GFX9-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX9-NEXT: s_mov_b64 s[0:1], exec -; GFX9-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX9-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX9-NEXT: .LBB3_1: ; %ComputeLoop ; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1 ; GFX9-NEXT: s_ff1_i32_b64 s2, s[0:1] @@ -1455,7 +1455,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1064-NEXT: s_mov_b32 s32, 0 ; GFX1064-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1064-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1064-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1064-NEXT: s_mov_b64 s[0:1], exec ; GFX1064-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1064-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1525,7 +1525,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1032-NEXT: s_mov_b32 s32, 0 ; GFX1032-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1032-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1032-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1032-NEXT: s_mov_b32 s0, exec_lo ; GFX1032-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1032-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1584,7 +1584,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1164-NEXT: s_mov_b32 s32, 0 ; GFX1164-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1164-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1164-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-NEXT: s_mov_b64 s[0:1], exec ; GFX1164-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1164-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1650,7 +1650,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1132-NEXT: s_mov_b32 s32, 0 ; GFX1132-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1132-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1132-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-NEXT: s_mov_b32 s0, exec_lo ; GFX1132-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1132-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1728,32 +1728,32 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX9-DPP-NEXT: v_mbcnt_hi_u32_b32 v1, exec_hi, v1 ; GFX9-DPP-NEXT: v_mov_b32_e32 v3, v0 ; GFX9-DPP-NEXT: s_not_b64 exec, exec -; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX9-DPP-NEXT: s_not_b64 exec, exec ; GFX9-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 -; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 0 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:1 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:2 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:4 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:8 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_bcast:15 row_mask:0xa bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 @@ -1818,25 +1818,25 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1064-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1064-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec -; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v4, v3 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -1910,25 +1910,25 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1032-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1032-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v4, v3 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -1984,30 +1984,30 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1164-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec -; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1164-DPP-NEXT: s_waitcnt_depctr 0xfff ; GFX1164-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v2, v2 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v2, v1 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -2079,26 +2079,26 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_one_as_scope ; GFX1132-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v2, v1 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -2565,7 +2565,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX9-NEXT: s_waitcnt lgkmcnt(0) ; GFX9-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX9-NEXT: s_mov_b64 s[0:1], exec -; GFX9-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX9-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX9-NEXT: .LBB5_1: ; %ComputeLoop ; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1 ; GFX9-NEXT: s_ff1_i32_b64 s2, s[0:1] @@ -2634,7 +2634,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1064-NEXT: s_mov_b32 s32, 0 ; GFX1064-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1064-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1064-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1064-NEXT: s_mov_b64 s[0:1], exec ; GFX1064-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1064-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2704,7 +2704,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1032-NEXT: s_mov_b32 s32, 0 ; GFX1032-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1032-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1032-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1032-NEXT: s_mov_b32 s0, exec_lo ; GFX1032-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1032-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2763,7 +2763,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1164-NEXT: s_mov_b32 s32, 0 ; GFX1164-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1164-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1164-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-NEXT: s_mov_b64 s[0:1], exec ; GFX1164-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1164-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2829,7 +2829,7 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1132-NEXT: s_mov_b32 s32, 0 ; GFX1132-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1132-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1132-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-NEXT: s_mov_b32 s0, exec_lo ; GFX1132-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1132-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2907,32 +2907,32 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX9-DPP-NEXT: v_mbcnt_hi_u32_b32 v1, exec_hi, v1 ; GFX9-DPP-NEXT: v_mov_b32_e32 v3, v0 ; GFX9-DPP-NEXT: s_not_b64 exec, exec -; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX9-DPP-NEXT: s_not_b64 exec, exec ; GFX9-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 -; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 0 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:1 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:2 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:4 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:8 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_bcast:15 row_mask:0xa bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 @@ -2997,25 +2997,25 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1064-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1064-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec -; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v4, v3 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -3089,25 +3089,25 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1032-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1032-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v4, v3 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0xff800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -3163,30 +3163,30 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1164-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec -; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1164-DPP-NEXT: s_waitcnt_depctr 0xfff ; GFX1164-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v2, v2 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v2, v1 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0xff800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -3258,26 +3258,26 @@ define amdgpu_kernel void @global_atomic_fmax_uni_address_div_value_default_scop ; GFX1132-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0xff800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0xff800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v2, v1 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0xff800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf diff --git a/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmin.ll b/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmin.ll index a9f49ad..60195ca 100644 --- a/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmin.ll +++ b/llvm/test/CodeGen/AMDGPU/global_atomics_scan_fmin.ll @@ -326,7 +326,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX9-NEXT: s_waitcnt lgkmcnt(0) ; GFX9-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX9-NEXT: s_mov_b64 s[0:1], exec -; GFX9-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX9-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX9-NEXT: .LBB1_1: ; %ComputeLoop ; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1 ; GFX9-NEXT: s_ff1_i32_b64 s2, s[0:1] @@ -395,7 +395,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1064-NEXT: s_mov_b32 s32, 0 ; GFX1064-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1064-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1064-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1064-NEXT: s_mov_b64 s[0:1], exec ; GFX1064-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1064-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -451,7 +451,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1032-NEXT: s_mov_b32 s32, 0 ; GFX1032-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1032-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1032-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1032-NEXT: s_mov_b32 s0, exec_lo ; GFX1032-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1032-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -496,7 +496,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1164-NEXT: s_mov_b32 s32, 0 ; GFX1164-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1164-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1164-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-NEXT: s_mov_b64 s[0:1], exec ; GFX1164-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1164-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -546,7 +546,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1132-NEXT: s_mov_b32 s32, 0 ; GFX1132-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1132-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1132-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-NEXT: s_mov_b32 s0, exec_lo ; GFX1132-NEXT: .LBB1_1: ; %ComputeLoop ; GFX1132-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -609,32 +609,32 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX9-DPP-NEXT: v_mbcnt_hi_u32_b32 v1, exec_hi, v1 ; GFX9-DPP-NEXT: v_mov_b32_e32 v3, v0 ; GFX9-DPP-NEXT: s_not_b64 exec, exec -; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX9-DPP-NEXT: s_not_b64 exec, exec ; GFX9-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 -; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 0 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:1 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:2 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:4 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:8 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_bcast:15 row_mask:0xa bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 @@ -699,25 +699,25 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1064-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1064-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec -; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v4, v3 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -777,25 +777,25 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1032-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1032-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v4, v3 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -837,30 +837,30 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1164-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec -; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1164-DPP-NEXT: s_waitcnt_depctr 0xfff ; GFX1164-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v2, v2 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v2, v1 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -916,26 +916,26 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_agent_scope_ ; GFX1132-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v2, v1 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -1386,7 +1386,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX9-NEXT: s_waitcnt lgkmcnt(0) ; GFX9-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX9-NEXT: s_mov_b64 s[0:1], exec -; GFX9-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX9-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX9-NEXT: .LBB3_1: ; %ComputeLoop ; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1 ; GFX9-NEXT: s_ff1_i32_b64 s2, s[0:1] @@ -1455,7 +1455,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1064-NEXT: s_mov_b32 s32, 0 ; GFX1064-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1064-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1064-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1064-NEXT: s_mov_b64 s[0:1], exec ; GFX1064-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1064-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1525,7 +1525,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1032-NEXT: s_mov_b32 s32, 0 ; GFX1032-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1032-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1032-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1032-NEXT: s_mov_b32 s0, exec_lo ; GFX1032-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1032-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1584,7 +1584,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1164-NEXT: s_mov_b32 s32, 0 ; GFX1164-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1164-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1164-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-NEXT: s_mov_b64 s[0:1], exec ; GFX1164-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1164-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1650,7 +1650,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1132-NEXT: s_mov_b32 s32, 0 ; GFX1132-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1132-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1132-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-NEXT: s_mov_b32 s0, exec_lo ; GFX1132-NEXT: .LBB3_1: ; %ComputeLoop ; GFX1132-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -1728,32 +1728,32 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX9-DPP-NEXT: v_mbcnt_hi_u32_b32 v1, exec_hi, v1 ; GFX9-DPP-NEXT: v_mov_b32_e32 v3, v0 ; GFX9-DPP-NEXT: s_not_b64 exec, exec -; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX9-DPP-NEXT: s_not_b64 exec, exec ; GFX9-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 -; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 0 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:1 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:2 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:4 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:8 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_bcast:15 row_mask:0xa bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 @@ -1818,25 +1818,25 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1064-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1064-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec -; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v4, v3 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -1910,25 +1910,25 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1032-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1032-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v4, v3 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -1984,30 +1984,30 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1164-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec -; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1164-DPP-NEXT: s_waitcnt_depctr 0xfff ; GFX1164-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v2, v2 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v2, v1 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -2079,26 +2079,26 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_one_as_scope ; GFX1132-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v2, v1 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -2565,7 +2565,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX9-NEXT: s_waitcnt lgkmcnt(0) ; GFX9-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX9-NEXT: s_mov_b64 s[0:1], exec -; GFX9-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX9-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX9-NEXT: .LBB5_1: ; %ComputeLoop ; GFX9-NEXT: ; =>This Inner Loop Header: Depth=1 ; GFX9-NEXT: s_ff1_i32_b64 s2, s[0:1] @@ -2634,7 +2634,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1064-NEXT: s_mov_b32 s32, 0 ; GFX1064-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1064-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1064-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1064-NEXT: s_mov_b64 s[0:1], exec ; GFX1064-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1064-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2704,7 +2704,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1032-NEXT: s_mov_b32 s32, 0 ; GFX1032-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-NEXT: s_swappc_b64 s[30:31], s[16:17] -; GFX1032-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1032-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1032-NEXT: s_mov_b32 s0, exec_lo ; GFX1032-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1032-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2763,7 +2763,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1164-NEXT: s_mov_b32 s32, 0 ; GFX1164-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1164-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1164-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-NEXT: s_mov_b64 s[0:1], exec ; GFX1164-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1164-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2829,7 +2829,7 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1132-NEXT: s_mov_b32 s32, 0 ; GFX1132-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-NEXT: s_swappc_b64 s[30:31], s[2:3] -; GFX1132-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1132-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-NEXT: s_mov_b32 s0, exec_lo ; GFX1132-NEXT: .LBB5_1: ; %ComputeLoop ; GFX1132-NEXT: ; =>This Inner Loop Header: Depth=1 @@ -2907,32 +2907,32 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX9-DPP-NEXT: v_mbcnt_hi_u32_b32 v1, exec_hi, v1 ; GFX9-DPP-NEXT: v_mov_b32_e32 v3, v0 ; GFX9-DPP-NEXT: s_not_b64 exec, exec -; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX9-DPP-NEXT: s_not_b64 exec, exec ; GFX9-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 -; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 0 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:1 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:2 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:4 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_shr:8 row_mask:0xf bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 ; GFX9-DPP-NEXT: v_min_f32_e32 v3, v3, v5 -; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX9-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX9-DPP-NEXT: s_nop 1 ; GFX9-DPP-NEXT: v_mov_b32_dpp v5, v3 row_bcast:15 row_mask:0xa bank_mask:0xf ; GFX9-DPP-NEXT: v_max_f32_e32 v5, v5, v5 @@ -2997,25 +2997,25 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1064-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1064-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1064-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec -; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1064-DPP-NEXT: s_not_b64 exec, exec ; GFX1064-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v4, v3 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1064-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1064-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1064-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1064-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -3089,25 +3089,25 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1032-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1032-DPP-NEXT: s_swappc_b64 s[30:31], s[16:17] ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1032-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, v0 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v4, 0x7fc00000 ; GFX1032-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1032-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v3, v4 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v4, v4 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_max_f32_e32 v3, v3, v3 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v4, v3 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 -; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7f800000 +; GFX1032-DPP-NEXT: v_mov_b32_e32 v5, 0x7fc00000 ; GFX1032-DPP-NEXT: v_min_f32_e32 v3, v3, v4 ; GFX1032-DPP-NEXT: v_mov_b32_dpp v5, v3 row_xmask:8 row_mask:0xf bank_mask:0xf ; GFX1032-DPP-NEXT: v_max_f32_e32 v4, v5, v5 @@ -3163,30 +3163,30 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1164-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1164-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1164-DPP-NEXT: s_mov_b64 exec, s[0:1] ; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec -; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1164-DPP-NEXT: s_not_b64 exec, exec ; GFX1164-DPP-NEXT: s_or_saveexec_b64 s[0:1], -1 ; GFX1164-DPP-NEXT: s_waitcnt_depctr 0xfff ; GFX1164-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v2, v2 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v2, v1 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1164-DPP-NEXT: v_max_f32_e32 v2, v3, v3 -; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7f800000 +; GFX1164-DPP-NEXT: v_mov_b32_e32 v3, 0x7fc00000 ; GFX1164-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1164-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1164-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf @@ -3258,26 +3258,26 @@ define amdgpu_kernel void @global_atomic_fmin_uni_address_div_value_default_scop ; GFX1132-DPP-NEXT: s_waitcnt lgkmcnt(0) ; GFX1132-DPP-NEXT: s_swappc_b64 s[30:31], s[2:3] ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 -; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7f800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v1, 0x7fc00000 ; GFX1132-DPP-NEXT: s_mov_b32 exec_lo, s0 ; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, v0 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo -; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7f800000 +; GFX1132-DPP-NEXT: v_mov_b32_e32 v2, 0x7fc00000 ; GFX1132-DPP-NEXT: s_not_b32 exec_lo, exec_lo ; GFX1132-DPP-NEXT: s_or_saveexec_b32 s0, -1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v1, v2 row_xmask:1 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v2, v2 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_max_f32_e32 v1, v1, v1 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v2, v1 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:2 row_mask:0xf bank_mask:0xf ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:4 row_mask:0xf bank_mask:0xf -; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7f800000 +; GFX1132-DPP-NEXT: v_dual_max_f32 v2, v3, v3 :: v_dual_mov_b32 v3, 0x7fc00000 ; GFX1132-DPP-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) ; GFX1132-DPP-NEXT: v_min_f32_e32 v1, v1, v2 ; GFX1132-DPP-NEXT: v_mov_b32_dpp v3, v1 row_xmask:8 row_mask:0xf bank_mask:0xf diff --git a/llvm/test/CodeGen/AMDGPU/trunc-store.ll b/llvm/test/CodeGen/AMDGPU/trunc-store.ll index e642a70..931953e 100644 --- a/llvm/test/CodeGen/AMDGPU/trunc-store.ll +++ b/llvm/test/CodeGen/AMDGPU/trunc-store.ll @@ -1,18 +1,422 @@ -; RUN: llc -mtriple=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s -; RUN: llc -mtriple=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 +; RUN: llc -mtriple=amdgcn -mcpu=verde < %s | FileCheck -check-prefixes=GCN,SI %s +; RUN: llc -mtriple=amdgcn -mcpu=tonga < %s | FileCheck -check-prefixes=GCN,VI %s -; FUNC-LABEL: {{^}}truncstore_arg_v16i32_to_v16i8: -; SI: buffer_store_dwordx4 define amdgpu_kernel void @truncstore_arg_v16i32_to_v16i8(ptr addrspace(1) %out, <16 x i32> %in) { +; SI-LABEL: truncstore_arg_v16i32_to_v16i8: +; SI: ; %bb.0: +; SI-NEXT: s_load_dwordx16 s[4:19], s[0:1], 0x19 +; SI-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x9 +; SI-NEXT: s_mov_b32 s3, 0xf000 +; SI-NEXT: s_mov_b32 s2, -1 +; SI-NEXT: s_waitcnt lgkmcnt(0) +; SI-NEXT: s_and_b32 s18, s18, 0xff +; SI-NEXT: s_lshl_b32 s17, s17, 8 +; SI-NEXT: s_and_b32 s16, s16, 0xff +; SI-NEXT: s_and_b32 s14, s14, 0xff +; SI-NEXT: s_lshl_b32 s13, s13, 8 +; SI-NEXT: s_and_b32 s12, s12, 0xff +; SI-NEXT: s_and_b32 s10, s10, 0xff +; SI-NEXT: s_lshl_b32 s9, s9, 8 +; SI-NEXT: s_and_b32 s8, s8, 0xff +; SI-NEXT: s_and_b32 s6, s6, 0xff +; SI-NEXT: s_lshl_b32 s5, s5, 8 +; SI-NEXT: s_and_b32 s4, s4, 0xff +; SI-NEXT: s_lshl_b32 s19, s19, 24 +; SI-NEXT: s_lshl_b32 s18, s18, 16 +; SI-NEXT: s_or_b32 s16, s16, s17 +; SI-NEXT: s_lshl_b32 s15, s15, 24 +; SI-NEXT: s_lshl_b32 s14, s14, 16 +; SI-NEXT: s_or_b32 s12, s12, s13 +; SI-NEXT: s_lshl_b32 s11, s11, 24 +; SI-NEXT: s_lshl_b32 s10, s10, 16 +; SI-NEXT: s_or_b32 s8, s8, s9 +; SI-NEXT: s_lshl_b32 s7, s7, 24 +; SI-NEXT: s_lshl_b32 s6, s6, 16 +; SI-NEXT: s_or_b32 s4, s4, s5 +; SI-NEXT: s_or_b32 s18, s19, s18 +; SI-NEXT: s_and_b32 s16, s16, 0xffff +; SI-NEXT: s_or_b32 s14, s15, s14 +; SI-NEXT: s_and_b32 s12, s12, 0xffff +; SI-NEXT: s_or_b32 s10, s11, s10 +; SI-NEXT: s_and_b32 s8, s8, 0xffff +; SI-NEXT: s_or_b32 s6, s7, s6 +; SI-NEXT: s_and_b32 s4, s4, 0xffff +; SI-NEXT: s_or_b32 s16, s16, s18 +; SI-NEXT: s_or_b32 s12, s12, s14 +; SI-NEXT: s_or_b32 s8, s8, s10 +; SI-NEXT: s_or_b32 s4, s4, s6 +; SI-NEXT: v_mov_b32_e32 v0, s4 +; SI-NEXT: v_mov_b32_e32 v1, s8 +; SI-NEXT: v_mov_b32_e32 v2, s12 +; SI-NEXT: v_mov_b32_e32 v3, s16 +; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0 +; SI-NEXT: s_endpgm +; +; VI-LABEL: truncstore_arg_v16i32_to_v16i8: +; VI: ; %bb.0: +; VI-NEXT: s_load_dwordx16 s[4:19], s[0:1], 0x64 +; VI-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24 +; VI-NEXT: s_waitcnt lgkmcnt(0) +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s17 +; VI-NEXT: v_mov_b32_e32 v1, s16 +; VI-NEXT: v_lshlrev_b16_e64 v2, 8, s19 +; VI-NEXT: v_mov_b32_e32 v3, s18 +; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v1, v3, v2 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v3, v0, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s13 +; VI-NEXT: v_mov_b32_e32 v1, s12 +; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v1, 8, s15 +; VI-NEXT: v_mov_b32_e32 v2, s14 +; VI-NEXT: v_or_b32_sdwa v1, v2, v1 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v2, v0, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s9 +; VI-NEXT: v_mov_b32_e32 v1, s8 +; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v1, 8, s11 +; VI-NEXT: v_mov_b32_e32 v4, s10 +; VI-NEXT: v_or_b32_sdwa v1, v4, v1 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v1, v0, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s5 +; VI-NEXT: v_mov_b32_e32 v4, s4 +; VI-NEXT: v_or_b32_sdwa v0, v4, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v4, 8, s7 +; VI-NEXT: v_mov_b32_e32 v5, s6 +; VI-NEXT: v_or_b32_sdwa v4, v5, v4 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v0, v0, v4 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_mov_b32_e32 v5, s1 +; VI-NEXT: v_mov_b32_e32 v4, s0 +; VI-NEXT: flat_store_dwordx4 v[4:5], v[0:3] +; VI-NEXT: s_endpgm %trunc = trunc <16 x i32> %in to <16 x i8> store <16 x i8> %trunc, ptr addrspace(1) %out ret void } -; FUNC-LABEL: {{^}}truncstore_arg_v16i64_to_v16i8: -; SI: buffer_store_dwordx4 define amdgpu_kernel void @truncstore_arg_v16i64_to_v16i8(ptr addrspace(1) %out, <16 x i64> %in) { +; SI-LABEL: truncstore_arg_v16i64_to_v16i8: +; SI: ; %bb.0: +; SI-NEXT: s_load_dwordx16 s[16:31], s[0:1], 0x39 +; SI-NEXT: s_load_dwordx2 s[36:37], s[0:1], 0x9 +; SI-NEXT: s_load_dwordx16 s[0:15], s[0:1], 0x29 +; SI-NEXT: s_mov_b32 s39, 0xf000 +; SI-NEXT: s_mov_b32 s38, -1 +; SI-NEXT: s_waitcnt lgkmcnt(0) +; SI-NEXT: s_and_b32 s3, s28, 0xff +; SI-NEXT: s_lshl_b32 s1, s30, 24 +; SI-NEXT: s_lshl_b32 s3, s3, 16 +; SI-NEXT: s_or_b32 s1, s1, s3 +; SI-NEXT: s_lshl_b32 s3, s26, 8 +; SI-NEXT: s_and_b32 s5, s24, 0xff +; SI-NEXT: s_or_b32 s3, s5, s3 +; SI-NEXT: s_and_b32 s3, s3, 0xffff +; SI-NEXT: s_and_b32 s5, s20, 0xff +; SI-NEXT: s_or_b32 s1, s3, s1 +; SI-NEXT: s_lshl_b32 s3, s22, 24 +; SI-NEXT: s_lshl_b32 s5, s5, 16 +; SI-NEXT: s_or_b32 s3, s3, s5 +; SI-NEXT: s_lshl_b32 s5, s18, 8 +; SI-NEXT: s_and_b32 s7, s16, 0xff +; SI-NEXT: s_or_b32 s5, s7, s5 +; SI-NEXT: s_and_b32 s5, s5, 0xffff +; SI-NEXT: s_and_b32 s7, s12, 0xff +; SI-NEXT: s_or_b32 s3, s5, s3 +; SI-NEXT: s_lshl_b32 s5, s14, 24 +; SI-NEXT: s_lshl_b32 s7, s7, 16 +; SI-NEXT: s_or_b32 s5, s5, s7 +; SI-NEXT: s_lshl_b32 s7, s10, 8 +; SI-NEXT: s_and_b32 s8, s8, 0xff +; SI-NEXT: s_and_b32 s4, s4, 0xff +; SI-NEXT: s_lshl_b32 s2, s2, 8 +; SI-NEXT: s_and_b32 s0, s0, 0xff +; SI-NEXT: s_or_b32 s7, s8, s7 +; SI-NEXT: s_lshl_b32 s6, s6, 24 +; SI-NEXT: s_lshl_b32 s4, s4, 16 +; SI-NEXT: s_or_b32 s0, s0, s2 +; SI-NEXT: s_and_b32 s7, s7, 0xffff +; SI-NEXT: s_or_b32 s4, s6, s4 +; SI-NEXT: s_and_b32 s0, s0, 0xffff +; SI-NEXT: s_or_b32 s5, s7, s5 +; SI-NEXT: s_or_b32 s0, s0, s4 +; SI-NEXT: v_mov_b32_e32 v0, s0 +; SI-NEXT: v_mov_b32_e32 v1, s5 +; SI-NEXT: v_mov_b32_e32 v2, s3 +; SI-NEXT: v_mov_b32_e32 v3, s1 +; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[36:39], 0 +; SI-NEXT: s_endpgm +; +; VI-LABEL: truncstore_arg_v16i64_to_v16i8: +; VI: ; %bb.0: +; VI-NEXT: s_load_dwordx16 s[16:31], s[0:1], 0xe4 +; VI-NEXT: s_load_dwordx2 s[34:35], s[0:1], 0x24 +; VI-NEXT: s_load_dwordx16 s[0:15], s[0:1], 0xa4 +; VI-NEXT: s_waitcnt lgkmcnt(0) +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s26 +; VI-NEXT: v_mov_b32_e32 v1, s24 +; VI-NEXT: v_lshlrev_b16_e64 v2, 8, s30 +; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_mov_b32_e32 v1, s28 +; VI-NEXT: v_or_b32_sdwa v1, v1, v2 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v3, v0, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s18 +; VI-NEXT: v_mov_b32_e32 v1, s16 +; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v1, 8, s22 +; VI-NEXT: v_mov_b32_e32 v2, s20 +; VI-NEXT: v_or_b32_sdwa v1, v2, v1 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v2, v0, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s10 +; VI-NEXT: v_mov_b32_e32 v1, s8 +; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v1, 8, s14 +; VI-NEXT: v_mov_b32_e32 v4, s12 +; VI-NEXT: v_or_b32_sdwa v1, v4, v1 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v1, v0, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v0, 8, s2 +; VI-NEXT: v_mov_b32_e32 v4, s0 +; VI-NEXT: v_or_b32_sdwa v0, v4, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e64 v4, 8, s6 +; VI-NEXT: v_mov_b32_e32 v5, s4 +; VI-NEXT: v_or_b32_sdwa v4, v5, v4 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v0, v0, v4 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_mov_b32_e32 v4, s34 +; VI-NEXT: v_mov_b32_e32 v5, s35 +; VI-NEXT: flat_store_dwordx4 v[4:5], v[0:3] +; VI-NEXT: s_endpgm %trunc = trunc <16 x i64> %in to <16 x i8> store <16 x i8> %trunc, ptr addrspace(1) %out ret void } + +define void @truncstore_v5i32_to_v5i1(ptr addrspace(1) %out, <5 x i32> %val) { +; SI-LABEL: truncstore_v5i32_to_v5i1: +; SI: ; %bb.0: +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SI-NEXT: v_and_b32_e32 v3, 1, v3 +; SI-NEXT: v_and_b32_e32 v2, 1, v2 +; SI-NEXT: v_lshlrev_b32_e32 v3, 1, v3 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_and_b32_e32 v3, 1, v4 +; SI-NEXT: v_lshlrev_b32_e32 v3, 2, v3 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_and_b32_e32 v3, 1, v5 +; SI-NEXT: v_lshlrev_b32_e32 v3, 3, v3 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_lshlrev_b32_e32 v3, 4, v6 +; SI-NEXT: s_mov_b32 s6, 0 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: s_mov_b32 s7, 0xf000 +; SI-NEXT: s_mov_b32 s4, s6 +; SI-NEXT: s_mov_b32 s5, s6 +; SI-NEXT: v_and_b32_e32 v2, 31, v2 +; SI-NEXT: buffer_store_byte v2, v[0:1], s[4:7], 0 addr64 +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) +; SI-NEXT: s_setpc_b64 s[30:31] +; +; VI-LABEL: truncstore_v5i32_to_v5i1: +; VI: ; %bb.0: +; VI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; VI-NEXT: v_and_b32_e32 v3, 1, v3 +; VI-NEXT: v_and_b32_e32 v2, 1, v2 +; VI-NEXT: v_lshlrev_b16_e32 v3, 1, v3 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_and_b32_e32 v3, 1, v4 +; VI-NEXT: v_lshlrev_b16_e32 v3, 2, v3 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_and_b32_e32 v3, 1, v5 +; VI-NEXT: v_lshlrev_b16_e32 v3, 3, v3 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_lshlrev_b16_e32 v3, 4, v6 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_and_b32_e32 v2, 31, v2 +; VI-NEXT: flat_store_byte v[0:1], v2 +; VI-NEXT: s_waitcnt vmcnt(0) +; VI-NEXT: s_setpc_b64 s[30:31] + %trunc = trunc <5 x i32> %val to <5 x i1> + store <5 x i1> %trunc, ptr addrspace(1) %out + ret void +} + +define void @truncstore_v5i32_to_v5i8(ptr addrspace(1) %out, <5 x i32> %val) { +; SI-LABEL: truncstore_v5i32_to_v5i8: +; SI: ; %bb.0: +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SI-NEXT: v_and_b32_e32 v4, 0xff, v4 +; SI-NEXT: v_lshlrev_b32_e32 v3, 8, v3 +; SI-NEXT: v_and_b32_e32 v2, 0xff, v2 +; SI-NEXT: s_mov_b32 s6, 0 +; SI-NEXT: v_lshlrev_b32_e32 v5, 24, v5 +; SI-NEXT: v_lshlrev_b32_e32 v4, 16, v4 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: s_mov_b32 s7, 0xf000 +; SI-NEXT: s_mov_b32 s4, s6 +; SI-NEXT: s_mov_b32 s5, s6 +; SI-NEXT: v_or_b32_e32 v4, v5, v4 +; SI-NEXT: v_and_b32_e32 v2, 0xffff, v2 +; SI-NEXT: v_or_b32_e32 v2, v2, v4 +; SI-NEXT: buffer_store_byte v6, v[0:1], s[4:7], 0 addr64 offset:4 +; SI-NEXT: buffer_store_dword v2, v[0:1], s[4:7], 0 addr64 +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) +; SI-NEXT: s_setpc_b64 s[30:31] +; +; VI-LABEL: truncstore_v5i32_to_v5i8: +; VI: ; %bb.0: +; VI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; VI-NEXT: v_lshlrev_b16_e32 v3, 8, v3 +; VI-NEXT: v_or_b32_sdwa v2, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e32 v3, 8, v5 +; VI-NEXT: v_or_b32_sdwa v3, v4, v3 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v4, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_add_u32_e32 v2, vcc, 4, v0 +; VI-NEXT: v_addc_u32_e32 v3, vcc, 0, v1, vcc +; VI-NEXT: flat_store_byte v[2:3], v6 +; VI-NEXT: flat_store_dword v[0:1], v4 +; VI-NEXT: s_waitcnt vmcnt(0) +; VI-NEXT: s_setpc_b64 s[30:31] + %trunc = trunc <5 x i32> %val to <5 x i8> + store <5 x i8> %trunc, ptr addrspace(1) %out + ret void +} + +define void @truncstore_v6i32_to_v6i1(ptr addrspace(1) %out, <6 x i32> %val) { +; SI-LABEL: truncstore_v6i32_to_v6i1: +; SI: ; %bb.0: +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SI-NEXT: v_and_b32_e32 v3, 1, v3 +; SI-NEXT: v_and_b32_e32 v2, 1, v2 +; SI-NEXT: v_lshlrev_b32_e32 v3, 1, v3 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_and_b32_e32 v3, 1, v4 +; SI-NEXT: v_lshlrev_b32_e32 v3, 2, v3 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_and_b32_e32 v3, 1, v5 +; SI-NEXT: v_lshlrev_b32_e32 v3, 3, v3 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_and_b32_e32 v3, 1, v6 +; SI-NEXT: v_lshlrev_b32_e32 v3, 4, v3 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_lshlrev_b32_e32 v3, 5, v7 +; SI-NEXT: s_mov_b32 s6, 0 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: s_mov_b32 s7, 0xf000 +; SI-NEXT: s_mov_b32 s4, s6 +; SI-NEXT: s_mov_b32 s5, s6 +; SI-NEXT: v_and_b32_e32 v2, 63, v2 +; SI-NEXT: buffer_store_byte v2, v[0:1], s[4:7], 0 addr64 +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) +; SI-NEXT: s_setpc_b64 s[30:31] +; +; VI-LABEL: truncstore_v6i32_to_v6i1: +; VI: ; %bb.0: +; VI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; VI-NEXT: v_and_b32_e32 v3, 1, v3 +; VI-NEXT: v_and_b32_e32 v2, 1, v2 +; VI-NEXT: v_lshlrev_b16_e32 v3, 1, v3 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_and_b32_e32 v3, 1, v4 +; VI-NEXT: v_lshlrev_b16_e32 v3, 2, v3 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_and_b32_e32 v3, 1, v5 +; VI-NEXT: v_lshlrev_b16_e32 v3, 3, v3 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_and_b32_e32 v3, 1, v6 +; VI-NEXT: v_lshlrev_b16_e32 v3, 4, v3 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_lshlrev_b16_e32 v3, 5, v7 +; VI-NEXT: v_or_b32_e32 v2, v2, v3 +; VI-NEXT: v_and_b32_e32 v2, 63, v2 +; VI-NEXT: flat_store_byte v[0:1], v2 +; VI-NEXT: s_waitcnt vmcnt(0) +; VI-NEXT: s_setpc_b64 s[30:31] + %trunc = trunc <6 x i32> %val to <6 x i1> + store <6 x i1> %trunc, ptr addrspace(1) %out + ret void +} + +define void @truncstore_v6i32_to_v6i8(ptr addrspace(1) %out, <6 x i32> %val) { +; SI-LABEL: truncstore_v6i32_to_v6i8: +; SI: ; %bb.0: +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SI-NEXT: v_and_b32_e32 v4, 0xff, v4 +; SI-NEXT: v_lshlrev_b32_e32 v3, 8, v3 +; SI-NEXT: v_and_b32_e32 v2, 0xff, v2 +; SI-NEXT: v_lshlrev_b32_e32 v5, 24, v5 +; SI-NEXT: v_lshlrev_b32_e32 v4, 16, v4 +; SI-NEXT: v_or_b32_e32 v2, v2, v3 +; SI-NEXT: v_or_b32_e32 v4, v5, v4 +; SI-NEXT: v_and_b32_e32 v2, 0xffff, v2 +; SI-NEXT: s_mov_b32 s6, 0 +; SI-NEXT: v_or_b32_e32 v2, v2, v4 +; SI-NEXT: v_lshlrev_b32_e32 v3, 8, v7 +; SI-NEXT: v_and_b32_e32 v4, 0xff, v6 +; SI-NEXT: s_mov_b32 s7, 0xf000 +; SI-NEXT: s_mov_b32 s4, s6 +; SI-NEXT: s_mov_b32 s5, s6 +; SI-NEXT: v_or_b32_e32 v3, v4, v3 +; SI-NEXT: buffer_store_short v3, v[0:1], s[4:7], 0 addr64 offset:4 +; SI-NEXT: buffer_store_dword v2, v[0:1], s[4:7], 0 addr64 +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) +; SI-NEXT: s_setpc_b64 s[30:31] +; +; VI-LABEL: truncstore_v6i32_to_v6i8: +; VI: ; %bb.0: +; VI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; VI-NEXT: v_lshlrev_b16_e32 v3, 8, v3 +; VI-NEXT: v_or_b32_sdwa v2, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e32 v3, 8, v5 +; VI-NEXT: v_or_b32_sdwa v3, v4, v3 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_or_b32_sdwa v4, v2, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; VI-NEXT: v_lshlrev_b16_e32 v2, 8, v7 +; VI-NEXT: v_or_b32_sdwa v5, v6, v2 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD +; VI-NEXT: v_add_u32_e32 v2, vcc, 4, v0 +; VI-NEXT: v_addc_u32_e32 v3, vcc, 0, v1, vcc +; VI-NEXT: flat_store_short v[2:3], v5 +; VI-NEXT: flat_store_dword v[0:1], v4 +; VI-NEXT: s_waitcnt vmcnt(0) +; VI-NEXT: s_setpc_b64 s[30:31] + %trunc = trunc <6 x i32> %val to <6 x i8> + store <6 x i8> %trunc, ptr addrspace(1) %out + ret void +} + +define void @truncstore_v6i32_to_v6i16(ptr addrspace(1) %out, <6 x i32> %val) { +; SI-LABEL: truncstore_v6i32_to_v6i16: +; SI: ; %bb.0: +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SI-NEXT: v_lshlrev_b32_e32 v5, 16, v5 +; SI-NEXT: v_and_b32_e32 v4, 0xffff, v4 +; SI-NEXT: v_lshlrev_b32_e32 v3, 16, v3 +; SI-NEXT: v_and_b32_e32 v2, 0xffff, v2 +; SI-NEXT: s_mov_b32 s6, 0 +; SI-NEXT: v_or_b32_e32 v4, v4, v5 +; SI-NEXT: v_or_b32_e32 v3, v2, v3 +; SI-NEXT: v_lshlrev_b32_e32 v2, 16, v7 +; SI-NEXT: v_and_b32_e32 v5, 0xffff, v6 +; SI-NEXT: s_mov_b32 s7, 0xf000 +; SI-NEXT: s_mov_b32 s4, s6 +; SI-NEXT: s_mov_b32 s5, s6 +; SI-NEXT: v_or_b32_e32 v2, v5, v2 +; SI-NEXT: buffer_store_dword v2, v[0:1], s[4:7], 0 addr64 offset:8 +; SI-NEXT: buffer_store_dwordx2 v[3:4], v[0:1], s[4:7], 0 addr64 +; SI-NEXT: s_waitcnt vmcnt(0) expcnt(0) +; SI-NEXT: s_setpc_b64 s[30:31] +; +; VI-LABEL: truncstore_v6i32_to_v6i16: +; VI: ; %bb.0: +; VI-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; VI-NEXT: s_mov_b32 s4, 0x1000504 +; VI-NEXT: v_perm_b32 v6, v6, v7, s4 +; VI-NEXT: v_perm_b32 v5, v4, v5, s4 +; VI-NEXT: v_perm_b32 v4, v2, v3, s4 +; VI-NEXT: flat_store_dwordx3 v[0:1], v[4:6] +; VI-NEXT: s_waitcnt vmcnt(0) +; VI-NEXT: s_setpc_b64 s[30:31] + %trunc = trunc <6 x i32> %val to <6 x i16> + store <6 x i16> %trunc, ptr addrspace(1) %out + ret void +} + +;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line: +; GCN: {{.*}} diff --git a/llvm/test/CodeGen/NVPTX/inline-asm-b128-test1.ll b/llvm/test/CodeGen/NVPTX/inline-asm-b128-test1.ll new file mode 100644 index 0000000..3232f40 --- /dev/null +++ b/llvm/test/CodeGen/NVPTX/inline-asm-b128-test1.ll @@ -0,0 +1,148 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --extra_scrub --version 5 +; RUN: llc < %s -march=nvptx -mcpu=sm_70 -mattr=+ptx83 | FileCheck %s +; RUN: %if ptxas %{ llc < %s -march=nvptx -mcpu=sm_70 -mattr=+ptx83 | %ptxas-verify -arch=sm_70 %} + +target triple = "nvptx64-nvidia-cuda" + +@value = internal addrspace(1) global i128 0, align 16 + +define void @test_b128_input_from_const() { +; CHECK-LABEL: test_b128_input_from_const( +; CHECK: { +; CHECK-NEXT: .reg .b32 %r<3>; +; CHECK-NEXT: .reg .b64 %rd<4>; +; CHECK-NEXT: .reg .b128 %rq<2>; +; CHECK-EMPTY: +; CHECK-NEXT: // %bb.0: +; CHECK-NEXT: mov.u64 %rd2, 0; +; CHECK-NEXT: mov.u64 %rd3, 42; +; CHECK-NEXT: mov.b128 %rq1, {%rd3, %rd2}; +; CHECK-NEXT: mov.u32 %r1, value; +; CHECK-NEXT: cvta.global.u32 %r2, %r1; +; CHECK-NEXT: cvt.u64.u32 %rd1, %r2; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { st.b128 [%rd1], %rq1; } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: ret; + + tail call void asm sideeffect "{ st.b128 [$0], $1; }", "l,q"(ptr nonnull addrspacecast (ptr addrspace(1) @value to ptr), i128 42) + ret void +} + +define void @test_b128_input_from_load(ptr nocapture readonly %data) { +; CHECK-LABEL: test_b128_input_from_load( +; CHECK: { +; CHECK-NEXT: .reg .b32 %r<5>; +; CHECK-NEXT: .reg .b64 %rd<4>; +; CHECK-NEXT: .reg .b128 %rq<2>; +; CHECK-EMPTY: +; CHECK-NEXT: // %bb.0: +; CHECK-NEXT: ld.param.u32 %r1, [test_b128_input_from_load_param_0]; +; CHECK-NEXT: cvta.to.global.u32 %r2, %r1; +; CHECK-NEXT: ld.global.u64 %rd2, [%r2+8]; +; CHECK-NEXT: ld.global.u64 %rd3, [%r2]; +; CHECK-NEXT: mov.b128 %rq1, {%rd3, %rd2}; +; CHECK-NEXT: mov.u32 %r3, value; +; CHECK-NEXT: cvta.global.u32 %r4, %r3; +; CHECK-NEXT: cvt.u64.u32 %rd1, %r4; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { st.b128 [%rd1], %rq1; } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: ret; + + %1 = addrspacecast ptr %data to ptr addrspace(1) + %2 = load <2 x i64>, ptr addrspace(1) %1, align 16 + %3 = bitcast <2 x i64> %2 to i128 + tail call void asm sideeffect "{ st.b128 [$0], $1; }", "l,q"(ptr nonnull addrspacecast (ptr addrspace(1) @value to ptr), i128 %3) + ret void +} + +define void @test_b128_input_from_select(ptr nocapture readonly %flag) { +; CHECK-LABEL: test_b128_input_from_select( +; CHECK: { +; CHECK-NEXT: .reg .pred %p<2>; +; CHECK-NEXT: .reg .b16 %rs<2>; +; CHECK-NEXT: .reg .b32 %r<5>; +; CHECK-NEXT: .reg .b64 %rd<4>; +; CHECK-NEXT: .reg .b128 %rq<2>; +; CHECK-EMPTY: +; CHECK-NEXT: // %bb.0: +; CHECK-NEXT: ld.param.u32 %r1, [test_b128_input_from_select_param_0]; +; CHECK-NEXT: cvta.to.global.u32 %r2, %r1; +; CHECK-NEXT: ld.global.u8 %rs1, [%r2]; +; CHECK-NEXT: setp.eq.s16 %p1, %rs1, 0; +; CHECK-NEXT: selp.b64 %rd2, 24, 42, %p1; +; CHECK-NEXT: mov.u64 %rd3, 0; +; CHECK-NEXT: mov.b128 %rq1, {%rd2, %rd3}; +; CHECK-NEXT: mov.u32 %r3, value; +; CHECK-NEXT: cvta.global.u32 %r4, %r3; +; CHECK-NEXT: cvt.u64.u32 %rd1, %r4; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { st.b128 [%rd1], %rq1; } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: ret; + + %1 = addrspacecast ptr %flag to ptr addrspace(1) + %2 = load i8, ptr addrspace(1) %1, align 1 + %3 = icmp eq i8 %2, 0 + %4 = select i1 %3, i128 24, i128 42 + tail call void asm sideeffect "{ st.b128 [$0], $1; }", "l,q"(ptr nonnull addrspacecast (ptr addrspace(1) @value to ptr), i128 %4) + ret void +} + +define void @test_store_b128_output() { +; CHECK-LABEL: test_store_b128_output( +; CHECK: { +; CHECK-NEXT: .reg .b64 %rd<5>; +; CHECK-NEXT: .reg .b128 %rq<2>; +; CHECK-EMPTY: +; CHECK-NEXT: // %bb.0: +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { mov.b128 %rq1, 41; } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: mov.b128 {%rd1, %rd2}, %rq1; +; CHECK-NEXT: add.cc.s64 %rd3, %rd1, 1; +; CHECK-NEXT: addc.cc.s64 %rd4, %rd2, 0; +; CHECK-NEXT: st.global.u64 [value+8], %rd4; +; CHECK-NEXT: st.global.u64 [value], %rd3; +; CHECK-NEXT: ret; + + %1 = tail call i128 asm "{ mov.b128 $0, 41; }", "=q"() + %add = add nsw i128 %1, 1 + %2 = bitcast i128 %add to <2 x i64> + store <2 x i64> %2, ptr addrspace(1) @value, align 16 + ret void +} + +define void @test_use_of_b128_output(ptr nocapture readonly %data) { +; CHECK-LABEL: test_use_of_b128_output( +; CHECK: { +; CHECK-NEXT: .reg .b32 %r<3>; +; CHECK-NEXT: .reg .b64 %rd<7>; +; CHECK-NEXT: .reg .b128 %rq<3>; +; CHECK-EMPTY: +; CHECK-NEXT: // %bb.0: +; CHECK-NEXT: ld.param.u32 %r1, [test_use_of_b128_output_param_0]; +; CHECK-NEXT: cvta.to.global.u32 %r2, %r1; +; CHECK-NEXT: ld.global.u64 %rd1, [%r2+8]; +; CHECK-NEXT: ld.global.u64 %rd2, [%r2]; +; CHECK-NEXT: mov.b128 %rq2, {%rd2, %rd1}; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { mov.b128 %rq1, %rq2; } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: mov.b128 {%rd3, %rd4}, %rq1; +; CHECK-NEXT: add.cc.s64 %rd5, %rd3, 1; +; CHECK-NEXT: addc.cc.s64 %rd6, %rd4, 0; +; CHECK-NEXT: st.global.u64 [value], %rd5; +; CHECK-NEXT: st.global.u64 [value+8], %rd6; +; CHECK-NEXT: ret; + + %1 = addrspacecast ptr %data to ptr addrspace(1) + %2 = load <2 x i64>, ptr addrspace(1) %1, align 16 + %3 = bitcast <2 x i64> %2 to i128 + %4 = tail call i128 asm "{ mov.b128 $0, $1; }", "=q,q"(i128 %3) + %add = add nsw i128 %4, 1 + %5 = bitcast i128 %add to <2 x i64> + store <2 x i64> %5, ptr addrspace(1) @value, align 16 + ret void +} diff --git a/llvm/test/CodeGen/NVPTX/inline-asm-b128-test2.ll b/llvm/test/CodeGen/NVPTX/inline-asm-b128-test2.ll new file mode 100644 index 0000000..3d1d7fb --- /dev/null +++ b/llvm/test/CodeGen/NVPTX/inline-asm-b128-test2.ll @@ -0,0 +1,122 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --extra_scrub --version 5 +; RUN: llc < %s -march=nvptx -mcpu=sm_70 -mattr=+ptx83 | FileCheck %s +; RUN: %if ptxas %{ llc < %s -march=nvptx -mcpu=sm_70 -mattr=+ptx83 | %ptxas-verify -arch=sm_70 %} + +target triple = "nvptx64-nvidia-cuda" + +@u128_max = internal addrspace(1) global i128 0, align 16 +@u128_zero = internal addrspace(1) global i128 0, align 16 +@i128_max = internal addrspace(1) global i128 0, align 16 +@i128_min = internal addrspace(1) global i128 0, align 16 +@v_u128_max = internal addrspace(1) global i128 0, align 16 +@v_u128_zero = internal addrspace(1) global i128 0, align 16 +@v_i128_max = internal addrspace(1) global i128 0, align 16 +@v_i128_min = internal addrspace(1) global i128 0, align 16 +@v64 = internal addrspace(1) global ptr null, align 8 + +define void @test_corner_values() { +; CHECK-LABEL: test_corner_values( +; CHECK: { +; CHECK-NEXT: .reg .b32 %r<20>; +; CHECK-NEXT: .reg .b64 %rd<17>; +; CHECK-NEXT: .reg .b128 %rq<5>; +; CHECK-EMPTY: +; CHECK-NEXT: // %bb.0: +; CHECK-NEXT: ld.global.u32 %r1, [v64]; +; CHECK-NEXT: add.s32 %r2, %r1, 8; +; CHECK-NEXT: mov.u64 %rd13, -1; +; CHECK-NEXT: mov.b128 %rq1, {%rd13, %rd13}; +; CHECK-NEXT: cvt.u64.u32 %rd1, %r1; +; CHECK-NEXT: cvt.u64.u32 %rd2, %r2; +; CHECK-NEXT: mov.u32 %r3, v_u128_max; +; CHECK-NEXT: cvta.global.u32 %r4, %r3; +; CHECK-NEXT: cvt.u64.u32 %rd3, %r4; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { +; CHECK-NEXT: .reg .b64 hi; +; CHECK-NEXT: .reg .b64 lo; +; CHECK-NEXT: mov.b128 {lo, hi}, %rq1; +; CHECK-NEXT: st.b64 [%rd1], lo; +; CHECK-NEXT: st.b64 [%rd2], hi; +; CHECK-NEXT: st.b128 [%rd3], %rq1; +; CHECK-NEXT: } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: ld.global.u32 %r5, [v64]; +; CHECK-NEXT: add.s32 %r6, %r5, 16; +; CHECK-NEXT: add.s32 %r7, %r5, 24; +; CHECK-NEXT: mov.u64 %rd14, 9223372036854775807; +; CHECK-NEXT: mov.b128 %rq2, {%rd13, %rd14}; +; CHECK-NEXT: mov.u32 %r8, v_i128_max; +; CHECK-NEXT: cvta.global.u32 %r9, %r8; +; CHECK-NEXT: cvt.u64.u32 %rd6, %r9; +; CHECK-NEXT: cvt.u64.u32 %rd4, %r6; +; CHECK-NEXT: cvt.u64.u32 %rd5, %r7; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { +; CHECK-NEXT: .reg .b64 hi; +; CHECK-NEXT: .reg .b64 lo; +; CHECK-NEXT: mov.b128 {lo, hi}, %rq2; +; CHECK-NEXT: st.b64 [%rd4], lo; +; CHECK-NEXT: st.b64 [%rd5], hi; +; CHECK-NEXT: st.b128 [%rd6], %rq2; +; CHECK-NEXT: } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: ld.global.u32 %r10, [v64]; +; CHECK-NEXT: add.s32 %r11, %r10, 32; +; CHECK-NEXT: add.s32 %r12, %r10, 40; +; CHECK-NEXT: mov.u64 %rd15, -9223372036854775808; +; CHECK-NEXT: mov.u64 %rd16, 0; +; CHECK-NEXT: mov.b128 %rq3, {%rd16, %rd15}; +; CHECK-NEXT: mov.u32 %r13, v_i128_min; +; CHECK-NEXT: cvta.global.u32 %r14, %r13; +; CHECK-NEXT: cvt.u64.u32 %rd9, %r14; +; CHECK-NEXT: cvt.u64.u32 %rd7, %r11; +; CHECK-NEXT: cvt.u64.u32 %rd8, %r12; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { +; CHECK-NEXT: .reg .b64 hi; +; CHECK-NEXT: .reg .b64 lo; +; CHECK-NEXT: mov.b128 {lo, hi}, %rq3; +; CHECK-NEXT: st.b64 [%rd7], lo; +; CHECK-NEXT: st.b64 [%rd8], hi; +; CHECK-NEXT: st.b128 [%rd9], %rq3; +; CHECK-NEXT: } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: ld.global.u32 %r15, [v64]; +; CHECK-NEXT: add.s32 %r16, %r15, 48; +; CHECK-NEXT: add.s32 %r17, %r15, 56; +; CHECK-NEXT: mov.b128 %rq4, {%rd16, %rd16}; +; CHECK-NEXT: mov.u32 %r18, v_u128_zero; +; CHECK-NEXT: cvta.global.u32 %r19, %r18; +; CHECK-NEXT: cvt.u64.u32 %rd12, %r19; +; CHECK-NEXT: cvt.u64.u32 %rd10, %r16; +; CHECK-NEXT: cvt.u64.u32 %rd11, %r17; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { +; CHECK-NEXT: .reg .b64 hi; +; CHECK-NEXT: .reg .b64 lo; +; CHECK-NEXT: mov.b128 {lo, hi}, %rq4; +; CHECK-NEXT: st.b64 [%rd10], lo; +; CHECK-NEXT: st.b64 [%rd11], hi; +; CHECK-NEXT: st.b128 [%rd12], %rq4; +; CHECK-NEXT: } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: ret; + + %1 = load ptr, ptr addrspace(1) @v64, align 8 + %2 = getelementptr inbounds i64, ptr %1, i64 1 + tail call void asm sideeffect "{\0A\09.reg .b64 hi;\0A\09.reg .b64 lo;\0A\09mov.b128 {lo, hi}, $0;\0A\09st.b64 [$1], lo;\0A\09st.b64 [$2], hi;\0A\09st.b128 [$3], $0;\0A\09}", "q,l,l,l"(i128 -1, ptr %1, ptr nonnull %2, ptr nonnull addrspacecast (ptr addrspace(1) @v_u128_max to ptr)) + %3 = load ptr, ptr addrspace(1) @v64, align 8 + %4 = getelementptr inbounds i64, ptr %3, i64 2 + %5 = getelementptr inbounds i64, ptr %3, i64 3 + tail call void asm sideeffect "{\0A\09.reg .b64 hi;\0A\09.reg .b64 lo;\0A\09mov.b128 {lo, hi}, $0;\0A\09st.b64 [$1], lo;\0A\09st.b64 [$2], hi;\0A\09st.b128 [$3], $0;\0A\09}", "q,l,l,l"(i128 170141183460469231731687303715884105727, ptr nonnull %4, ptr nonnull %5, ptr nonnull addrspacecast (ptr addrspace(1) @v_i128_max to ptr)) + %6 = load ptr, ptr addrspace(1) @v64, align 8 + %7 = getelementptr inbounds i64, ptr %6, i64 4 + %8 = getelementptr inbounds i64, ptr %6, i64 5 + tail call void asm sideeffect "{\0A\09.reg .b64 hi;\0A\09.reg .b64 lo;\0A\09mov.b128 {lo, hi}, $0;\0A\09st.b64 [$1], lo;\0A\09st.b64 [$2], hi;\0A\09st.b128 [$3], $0;\0A\09}", "q,l,l,l"(i128 -170141183460469231731687303715884105728, ptr nonnull %7, ptr nonnull %8, ptr nonnull addrspacecast (ptr addrspace(1) @v_i128_min to ptr)) + %9 = load ptr, ptr addrspace(1) @v64, align 8 + %10 = getelementptr inbounds i64, ptr %9, i64 6 + %11 = getelementptr inbounds i64, ptr %9, i64 7 + tail call void asm sideeffect "{\0A\09.reg .b64 hi;\0A\09.reg .b64 lo;\0A\09mov.b128 {lo, hi}, $0;\0A\09st.b64 [$1], lo;\0A\09st.b64 [$2], hi;\0A\09st.b128 [$3], $0;\0A\09}", "q,l,l,l"(i128 0, ptr nonnull %10, ptr nonnull %11, ptr nonnull addrspacecast (ptr addrspace(1) @v_u128_zero to ptr)) + ret void +} diff --git a/llvm/test/CodeGen/NVPTX/inline-asm-b128-test3.ll b/llvm/test/CodeGen/NVPTX/inline-asm-b128-test3.ll new file mode 100644 index 0000000..ae45397 --- /dev/null +++ b/llvm/test/CodeGen/NVPTX/inline-asm-b128-test3.ll @@ -0,0 +1,67 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --extra_scrub --version 5 +; RUN: llc < %s -march=nvptx -mcpu=sm_70 -mattr=+ptx83 | FileCheck %s +; RUN: %if ptxas %{ llc < %s -march=nvptx -mcpu=sm_70 -mattr=+ptx83 | %ptxas-verify -arch=sm_70 %} + +target triple = "nvptx64-nvidia-cuda" + +@size = internal addrspace(1) global i32 0, align 4 +@x = internal addrspace(1) global i128 0, align 16 + +define void @test_b128_in_loop() { +; CHECK-LABEL: test_b128_in_loop( +; CHECK: { +; CHECK-NEXT: .reg .pred %p<3>; +; CHECK-NEXT: .reg .b64 %rd<15>; +; CHECK-NEXT: .reg .b128 %rq<3>; +; CHECK-EMPTY: +; CHECK-NEXT: // %bb.0: +; CHECK-NEXT: ld.global.s32 %rd1, [size]; +; CHECK-NEXT: setp.eq.s64 %p1, %rd1, 0; +; CHECK-NEXT: @%p1 bra $L__BB0_3; +; CHECK-NEXT: // %bb.1: // %BB1 +; CHECK-NEXT: ld.global.u64 %rd13, [x+8]; +; CHECK-NEXT: ld.global.u64 %rd12, [x]; +; CHECK-NEXT: mov.u64 %rd14, 0; +; CHECK-NEXT: $L__BB0_2: // %BB2 +; CHECK-NEXT: // =>This Inner Loop Header: Depth=1 +; CHECK-NEXT: mov.b128 %rq1, {%rd12, %rd13}; +; CHECK-NEXT: // begin inline asm +; CHECK-NEXT: { +; CHECK-NEXT: .reg .b64 lo; +; CHECK-NEXT: .reg .b64 hi; +; CHECK-NEXT: mov.b128 {lo, hi}, %rq1; +; CHECK-NEXT: add.cc.u64 lo, lo, %rd14; +; CHECK-NEXT: mov.b128 %rq1, {lo, hi}; +; CHECK-NEXT: } +; CHECK-NEXT: // end inline asm +; CHECK-NEXT: mov.b128 {%rd12, %rd13}, %rq1; +; CHECK-NEXT: st.global.u64 [x+8], %rd13; +; CHECK-NEXT: st.global.u64 [x], %rd12; +; CHECK-NEXT: add.s64 %rd14, %rd14, 1; +; CHECK-NEXT: setp.ne.s64 %p2, %rd1, %rd14; +; CHECK-NEXT: @%p2 bra $L__BB0_2; +; CHECK-NEXT: $L__BB0_3: // %BB3 +; CHECK-NEXT: ret; + + %1 = load i32, ptr addrspace(1) @size, align 4 + %2 = icmp eq i32 %1, 0 + br i1 %2, label %BB3, label %BB1 + +BB1: ; preds = %0 + %3 = load i128, ptr addrspace(1) @x, align 16 + %4 = sext i32 %1 to i64 + br label %BB2 + +BB2: ; preds = %BB2, %BB1 + %5 = phi i128 [ %7, %BB2 ], [ %3, %BB1 ] + %6 = phi i64 [ %9, %BB2 ], [ 0, %BB1 ] + %7 = tail call i128 asm "{\0A\09.reg .b64 lo;\0A\09.reg .b64 hi;\0A\09mov.b128 {lo, hi}, $0;\0A\09add.cc.u64 lo, lo, $1;\0A\09mov.b128 $0, {lo, hi};\0A\09}", "=q,l,0"(i64 %6, i128 %5) + %8 = bitcast i128 %7 to <2 x i64> + store <2 x i64> %8, ptr addrspace(1) @x, align 16 + %9 = add nuw i64 %6, 1 + %10 = icmp eq i64 %9, %4 + br i1 %10, label %BB3, label %BB2 + +BB3: ; preds = %BB2, %0 + ret void +} diff --git a/llvm/test/CodeGen/RISCV/O3-pipeline.ll b/llvm/test/CodeGen/RISCV/O3-pipeline.ll index 1eee62e..3611d92 100644 --- a/llvm/test/CodeGen/RISCV/O3-pipeline.ll +++ b/llvm/test/CodeGen/RISCV/O3-pipeline.ll @@ -182,9 +182,9 @@ ; CHECK-NEXT: Insert fentry calls ; CHECK-NEXT: Insert XRay ops ; CHECK-NEXT: Implement the 'patchable-function' attribute +; CHECK-NEXT: Machine Copy Propagation Pass ; CHECK-NEXT: Branch relaxation pass ; CHECK-NEXT: RISC-V Make Compressible -; CHECK-NEXT: Machine Copy Propagation Pass ; CHECK-NEXT: Contiguously Lay Out Funclets ; CHECK-NEXT: StackMap Liveness Analysis ; CHECK-NEXT: Live DEBUG_VALUE analysis diff --git a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll index 68e8f5d..884c756 100644 --- a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll +++ b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll @@ -1062,3 +1062,36 @@ exit: %c = call <vscale x 2 x i32> @llvm.riscv.vadd.nxv2i32(<vscale x 2 x i32> undef, <vscale x 2 x i32> %a, <vscale x 2 x i32> %d, i64 %vl) ret <vscale x 2 x i32> %c } + +define void @vlmax_avl_phi(i1 %cmp, ptr %p, i64 %a, i64 %b) { +; CHECK-LABEL: vlmax_avl_phi: +; CHECK: # %bb.0: # %entry +; CHECK-NEXT: andi a0, a0, 1 +; CHECK-NEXT: beqz a0, .LBB25_2 +; CHECK-NEXT: # %bb.1: # %foo +; CHECK-NEXT: vsetvli zero, a2, e8, m1, ta, ma +; CHECK-NEXT: j .LBB25_3 +; CHECK-NEXT: .LBB25_2: # %bar +; CHECK-NEXT: vsetvli zero, a3, e8, m1, ta, ma +; CHECK-NEXT: .LBB25_3: # %exit +; CHECK-NEXT: vmv.v.i v8, 0 +; CHECK-NEXT: vsetivli zero, 1, e8, m1, ta, ma +; CHECK-NEXT: vse8.v v8, (a1) +; CHECK-NEXT: ret +entry: + br i1 %cmp, label %foo, label %bar + +foo: + %vl.foo = tail call i64 @llvm.riscv.vsetvli.i64(i64 %a, i64 0, i64 0) + br label %exit + +bar: + %vl.bar = tail call i64 @llvm.riscv.vsetvli.i64(i64 %b, i64 0, i64 0) + br label %exit + +exit: + %phivl = phi i64 [ %vl.foo, %foo ], [ %vl.bar, %bar ] + %1 = tail call <vscale x 8 x i8> @llvm.riscv.vmv.v.x.nxv8i8.i64(<vscale x 8 x i8> poison, i8 0, i64 %phivl) + call void @llvm.riscv.vse.nxv8i8(<vscale x 8 x i8> %1, ptr %p, i64 1) + ret void +} diff --git a/llvm/test/CodeGen/WebAssembly/call-indirect.ll b/llvm/test/CodeGen/WebAssembly/call-indirect.ll index b88d096..55a654f 100644 --- a/llvm/test/CodeGen/WebAssembly/call-indirect.ll +++ b/llvm/test/CodeGen/WebAssembly/call-indirect.ll @@ -1,4 +1,4 @@ -; RUN: llc < %s -asm-verbose=false -O2 | FileCheck --check-prefixes=CHECK,NOREF %s +; RUN: llc < %s -asm-verbose=false -mattr=-reference-types -O2 | FileCheck --check-prefixes=CHECK,NOREF %s ; RUN: llc < %s -asm-verbose=false -mattr=+reference-types -O2 | FileCheck --check-prefixes=CHECK,REF %s ; RUN: llc < %s -asm-verbose=false -O2 --filetype=obj | obj2yaml | FileCheck --check-prefix=OBJ %s diff --git a/llvm/test/CodeGen/WebAssembly/function-pointer64.ll b/llvm/test/CodeGen/WebAssembly/function-pointer64.ll index 7f98d3e..d5d10b0 100644 --- a/llvm/test/CodeGen/WebAssembly/function-pointer64.ll +++ b/llvm/test/CodeGen/WebAssembly/function-pointer64.ll @@ -1,6 +1,6 @@ -; RUN: llc < %s -asm-verbose=false -O2 | FileCheck %s +; RUN: llc < %s -asm-verbose=false -mattr=-reference-types -O2 | FileCheck %s ; RUN: llc < %s -asm-verbose=false -mattr=+reference-types -O2 | FileCheck --check-prefix=REF %s -; RUN: llc < %s -asm-verbose=false -O2 --filetype=obj | obj2yaml | FileCheck --check-prefix=YAML %s +; RUN: llc < %s -asm-verbose=false -mattr=-reference-types -O2 --filetype=obj | obj2yaml | FileCheck --check-prefix=YAML %s ; This tests pointer features that may codegen differently in wasm64. diff --git a/llvm/test/CodeGen/WebAssembly/reg-stackify.ll b/llvm/test/CodeGen/WebAssembly/reg-stackify.ll index f9845d4..f1466a4 100644 --- a/llvm/test/CodeGen/WebAssembly/reg-stackify.ll +++ b/llvm/test/CodeGen/WebAssembly/reg-stackify.ll @@ -649,7 +649,7 @@ define i32 @stackpointer_dependency(ptr readnone) { ; NOREGS-NEXT: local.tee 0 ; NOREGS: i32.load 0 ; NOREGS-NEXT: i32.load 0 -; NOREGS-NEXT: call_indirect (i32, i32) -> (i32) +; NOREGS-NEXT: call_indirect __indirect_function_table, (i32, i32) -> (i32) %class.call_indirect = type { ptr } define i32 @call_indirect_stackify(ptr %objptr, i32 %arg) { %obj = load ptr, ptr %objptr diff --git a/llvm/test/CodeGen/WebAssembly/swiftcc.ll b/llvm/test/CodeGen/WebAssembly/swiftcc.ll index e0f67b0..0c5c3d8 100644 --- a/llvm/test/CodeGen/WebAssembly/swiftcc.ll +++ b/llvm/test/CodeGen/WebAssembly/swiftcc.ll @@ -19,21 +19,21 @@ define swiftcc void @bar() { call swiftcc void @foo(i32 1, i32 2) ; REG: call_indirect $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}} -; CHECK: call_indirect (i32, i32, i32, i32) -> () +; CHECK: call_indirect __indirect_function_table, (i32, i32, i32, i32) -> () call swiftcc void %1(i32 1, i32 2) ; REG: call_indirect $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}} -; CHECK: call_indirect (i32, i32, i32, i32) -> () +; CHECK: call_indirect __indirect_function_table, (i32, i32, i32, i32) -> () call swiftcc void %1(i32 1, i32 2, i32 swiftself 3) %err = alloca swifterror ptr, align 4 ; REG: call_indirect $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}} -; CHECK: call_indirect (i32, i32, i32, i32) -> () +; CHECK: call_indirect __indirect_function_table, (i32, i32, i32, i32) -> () call swiftcc void %1(i32 1, i32 2, ptr swifterror %err) ; REG: call_indirect $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}}, $pop{{[0-9]+}} -; CHECK: call_indirect (i32, i32, i32, i32) -> () +; CHECK: call_indirect __indirect_function_table, (i32, i32, i32, i32) -> () call swiftcc void %1(i32 1, i32 2, i32 swiftself 3, ptr swifterror %err) ret void diff --git a/llvm/test/CodeGen/WebAssembly/target-features-cpus.ll b/llvm/test/CodeGen/WebAssembly/target-features-cpus.ll index d368271..d931475 100644 --- a/llvm/test/CodeGen/WebAssembly/target-features-cpus.ll +++ b/llvm/test/CodeGen/WebAssembly/target-features-cpus.ll @@ -11,20 +11,28 @@ target triple = "wasm32-unknown-unknown" ; mvp: should not contain the target features section ; MVP-NOT: .custom_section.target_features,"",@ -; generic: +mutable-globals, +sign-ext +; generic: +multivalue, +mutable-globals, +reference-types, +sign-ext ; GENERIC-LABEL: .custom_section.target_features,"",@ -; GENERIC-NEXT: .int8 2 +; GENERIC-NEXT: .int8 4 +; GENERIC-NEXT: .int8 43 +; GENERIC-NEXT: .int8 10 +; GENERIC-NEXT: .ascii "multivalue" ; GENERIC-NEXT: .int8 43 ; GENERIC-NEXT: .int8 15 ; GENERIC-NEXT: .ascii "mutable-globals" ; GENERIC-NEXT: .int8 43 +; GENERIC-NEXT: .int8 15 +; GENERIC-NEXT: .ascii "reference-types" +; GENERIC-NEXT: .int8 43 ; GENERIC-NEXT: .int8 8 ; GENERIC-NEXT: .ascii "sign-ext" -; bleeding-edge: +atomics, +bulk-memory, +mutable-globals, +nontrapping-fptoint, -; +sign-ext, +simd128, +tail-call +; bleeding-edge: +atomics, +bulk-memory, +exception-handling, +extended-const, +; +half-precision, +multimemory, +multivalue, +mutable-globals, +; +nontrapping-fptoint, +relaxed-simd, +reference-types, +; +simd128, +sign-ext, +tail-call ; BLEEDING-EDGE-LABEL: .section .custom_section.target_features,"",@ -; BLEEDING-EDGE-NEXT: .int8 7 +; BLEEDING-EDGE-NEXT: .int8 14 ; BLEEDING-EDGE-NEXT: .int8 43 ; BLEEDING-EDGE-NEXT: .int8 7 ; BLEEDING-EDGE-NEXT: .ascii "atomics" @@ -32,12 +40,33 @@ target triple = "wasm32-unknown-unknown" ; BLEEDING-EDGE-NEXT: .int8 11 ; BLEEDING-EDGE-NEXT: .ascii "bulk-memory" ; BLEEDING-EDGE-NEXT: .int8 43 +; BLEEDING-EDGE-NEXT: .int8 18 +; BLEEDING-EDGE-NEXT: .ascii "exception-handling" +; BLEEDING-EDGE-NEXT: .int8 43 +; BLEEDING-EDGE-NEXT: .int8 14 +; BLEEDING-EDGE-NEXT: .ascii "extended-const" +; BLEEDING-EDGE-NEXT: .int8 43 +; BLEEDING-EDGE-NEXT: .int8 14 +; BLEEDING-EDGE-NEXT: .ascii "half-precision" +; BLEEDING-EDGE-NEXT: .int8 43 +; BLEEDING-EDGE-NEXT: .int8 11 +; BLEEDING-EDGE-NEXT: .ascii "multimemory" +; BLEEDING-EDGE-NEXT: .int8 43 +; BLEEDING-EDGE-NEXT: .int8 10 +; BLEEDING-EDGE-NEXT: .ascii "multivalue" +; BLEEDING-EDGE-NEXT: .int8 43 ; BLEEDING-EDGE-NEXT: .int8 15 ; BLEEDING-EDGE-NEXT: .ascii "mutable-globals" ; BLEEDING-EDGE-NEXT: .int8 43 ; BLEEDING-EDGE-NEXT: .int8 19 ; BLEEDING-EDGE-NEXT: .ascii "nontrapping-fptoint" ; BLEEDING-EDGE-NEXT: .int8 43 +; BLEEDING-EDGE-NEXT: .int8 15 +; BLEEDING-EDGE-NEXT: .ascii "reference-types" +; BLEEDING-EDGE-NEXT: .int8 43 +; BLEEDING-EDGE-NEXT: .int8 12 +; BLEEDING-EDGE-NEXT: .ascii "relaxed-simd" +; BLEEDING-EDGE-NEXT: .int8 43 ; BLEEDING-EDGE-NEXT: .int8 8 ; BLEEDING-EDGE-NEXT: .ascii "sign-ext" ; BLEEDING-EDGE-NEXT: .int8 43 diff --git a/llvm/test/MC/WebAssembly/function-alias.ll b/llvm/test/MC/WebAssembly/function-alias.ll index 6722d18..036cd7d 100644 --- a/llvm/test/MC/WebAssembly/function-alias.ll +++ b/llvm/test/MC/WebAssembly/function-alias.ll @@ -1,4 +1,4 @@ -; RUN: llc -filetype=obj %s -o - | llvm-readobj --symbols - | FileCheck %s +; RUN: llc -filetype=obj %s -mattr=-reference-types -o - | llvm-readobj --symbols - | FileCheck %s ; RUN: llc -filetype=obj %s -mattr=+reference-types -o - | llvm-readobj --symbols - | FileCheck --check-prefix=REF %s target triple = "wasm32-unknown-unknown-wasm" diff --git a/llvm/test/MC/WebAssembly/no-dead-strip.ll b/llvm/test/MC/WebAssembly/no-dead-strip.ll index 6b3f090..3125dcf 100644 --- a/llvm/test/MC/WebAssembly/no-dead-strip.ll +++ b/llvm/test/MC/WebAssembly/no-dead-strip.ll @@ -50,6 +50,11 @@ entry: ; CHECK-NEXT: Segment: 3 ; CHECK-NEXT: Offset: 8 ; CHECK-NEXT: Size: 8 +; CHECK-NEXT: - Index: 5 +; CHECK-NEXT: Kind: TABLE +; CHECK-NEXT: Name: __indirect_function_table +; CHECK-NEXT: Flags: [ UNDEFINED, NO_STRIP ] +; CHECK-NEXT: Table: 0 ; CHECK-NEXT: SegmentInfo: ; CHECK-NEXT: - Index: 0 ; CHECK-NEXT: Name: .data.gv0 diff --git a/llvm/test/Transforms/Attributor/phi_bug_pointer_info.ll b/llvm/test/Transforms/Attributor/phi_bug_pointer_info.ll new file mode 100644 index 0000000..bb423e1 --- /dev/null +++ b/llvm/test/Transforms/Attributor/phi_bug_pointer_info.ll @@ -0,0 +1,41 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-attributes --check-globals --version 2 +; RUN: opt -aa-pipeline=basic-aa -passes=attributor -debug-only=attributor -attributor-manifest-internal -attributor-annotate-decl-cs -S < %s 2>&1 | FileCheck %s +; RUN: opt -aa-pipeline=basic-aa -passes=attributor-cgscc -debug-only=attributor -attributor-manifest-internal -attributor-annotate-decl-cs -S < %s 2>&1 | FileCheck %s +; REQUIRES: asserts + + +@globalBytes = internal global [1024 x i8] zeroinitializer + +; CHECK: Accesses by bin after update: +; CHECK: [8-12] : 2 +; CHECK: - 9 - store i32 %0, ptr %field2, align 4 +; CHECK: - c: %0 = load i32, ptr %val, align 4 +; CHECK: - 6 - %ret = load i32, ptr %x, align 4 +; CHECK: - c: <unknown> +; CHECK: [32-36] : 2 +; CHECK: - 9 - store i32 %1, ptr %field8, align 4 +; CHECK: - c: %1 = load i32, ptr %val2, align 4 +; CHECK: - 6 - %ret = load i32, ptr %x, align 4 +; CHECK: - c: <unknown> +define dso_local i32 @phi_different_offsets(ptr nocapture %val, ptr nocapture %val2, i1 %cmp) { +entry: + br i1 %cmp, label %then, label %else + +then: + %field2 = getelementptr i32, ptr @globalBytes, i32 2 + %1 = load i32, ptr %val + store i32 %1, ptr %field2 + br label %end + +else: + %field8 = getelementptr i32, ptr @globalBytes, i32 8 + %2 = load i32, ptr %val2 + store i32 %2, ptr %field8 + br label %end + +end: + %x = phi ptr [ %field2, %then ], [ %field8, %else ] + %ret = load i32, ptr %x + ret i32 %ret + +} diff --git a/llvm/test/Transforms/CorrelatedValuePropagation/uscmp.ll b/llvm/test/Transforms/CorrelatedValuePropagation/uscmp.ll new file mode 100644 index 0000000..efe4235 --- /dev/null +++ b/llvm/test/Transforms/CorrelatedValuePropagation/uscmp.ll @@ -0,0 +1,258 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; RUN: opt < %s -passes=correlated-propagation -S | FileCheck %s + +; If nothing is known we can't change anything +define i8 @ucmp_0(i32 %x, i32 %y) { +; CHECK-LABEL: @ucmp_0( +; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.ucmp.i8.i32(i32 [[X:%.*]], i32 [[Y:%.*]]) +; CHECK-NEXT: ret i8 [[TMP1]] +; + %1 = call i8 @llvm.ucmp(i32 %x, i32 %y) + ret i8 %1 +} + +define i8 @scmp_0(i32 %x, i32 %y) { +; CHECK-LABEL: @scmp_0( +; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.scmp.i8.i32(i32 [[X:%.*]], i32 [[Y:%.*]]) +; CHECK-NEXT: ret i8 [[TMP1]] +; + %1 = call i8 @llvm.scmp(i32 %x, i32 %y) + ret i8 %1 +} + +; If we know that range of LHS < range of RHS then return -1 +define i8 @ucmp_1(i32 %x, i32 %y) { + ; X is within [4, 8) +; CHECK-LABEL: @ucmp_1( +; CHECK-NEXT: [[COND1:%.*]] = icmp uge i32 [[X:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp ult i32 [[X]], 8 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: [[COND3:%.*]] = icmp uge i32 [[Y:%.*]], 8 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND3]]) +; CHECK-NEXT: ret i8 -1 +; + %cond1 = icmp uge i32 %x, 4 + call void @llvm.assume(i1 %cond1) + %cond2 = icmp ult i32 %x, 8 + call void @llvm.assume(i1 %cond2) + ; Y is within [8, UNSIGNED_MAX) + %cond3 = icmp uge i32 %y, 8 + call void @llvm.assume(i1 %cond3) + + %1 = call i8 @llvm.ucmp(i32 %x, i32 %y) + ret i8 %1 +} + +define i8 @scmp_1(i32 %x, i32 %y) { + ; X is within [-5, 3) +; CHECK-LABEL: @scmp_1( +; CHECK-NEXT: [[COND1:%.*]] = icmp sge i32 [[X:%.*]], -5 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp slt i32 [[X]], 3 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: [[COND3:%.*]] = icmp sge i32 [[Y:%.*]], 3 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND3]]) +; CHECK-NEXT: ret i8 -1 +; + %cond1 = icmp sge i32 %x, -5 + call void @llvm.assume(i1 %cond1) + %cond2 = icmp slt i32 %x, 3 + call void @llvm.assume(i1 %cond2) + ; Y is within [3, SIGNED_MAX) + %cond3 = icmp sge i32 %y, 3 + call void @llvm.assume(i1 %cond3) + + %1 = call i8 @llvm.scmp(i32 %x, i32 %y) + ret i8 %1 +} + +; If we know that range of LHS > range of RHS then return 1 +define i8 @ucmp_2(i32 %x, i32 %y) { + ; X is within [4, UNSIGNED_MAX) +; CHECK-LABEL: @ucmp_2( +; CHECK-NEXT: [[COND1:%.*]] = icmp uge i32 [[X:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp ult i32 [[Y:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: ret i8 1 +; + %cond1 = icmp uge i32 %x, 4 + call void @llvm.assume(i1 %cond1) + ; Y is within [0, 4) + %cond2 = icmp ult i32 %y, 4 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.ucmp(i32 %x, i32 %y) + ret i8 %1 +} + +define i8 @scmp_2(i32 %x, i32 %y) { + ; X is within [4, SIGNED_MAX) +; CHECK-LABEL: @scmp_2( +; CHECK-NEXT: [[COND1:%.*]] = icmp sge i32 [[X:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp slt i32 [[Y:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: ret i8 1 +; + %cond1 = icmp sge i32 %x, 4 + call void @llvm.assume(i1 %cond1) + ; Y is within [SIGNED_MIN, 4) + %cond2 = icmp slt i32 %y, 4 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.scmp(i32 %x, i32 %y) + ret i8 %1 +} + +; If we know that LHS and RHS are both constants then return 0 +define i8 @ucmp_5(i32 %x, i32 %y) { +; CHECK-LABEL: @ucmp_5( +; CHECK-NEXT: [[COND1:%.*]] = icmp eq i32 [[X:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp eq i32 [[Y:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: ret i8 0 +; + %cond1 = icmp eq i32 %x, 4 + call void @llvm.assume(i1 %cond1) + %cond2 = icmp eq i32 %y, 4 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.ucmp(i32 %x, i32 %y) + ret i8 %1 +} + +define i8 @scmp_5(i32 %x, i32 %y) { +; CHECK-LABEL: @scmp_5( +; CHECK-NEXT: [[COND1:%.*]] = icmp eq i32 [[X:%.*]], -5 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp eq i32 [[Y:%.*]], -5 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: ret i8 0 +; + %cond1 = icmp eq i32 %x, -5 + call void @llvm.assume(i1 %cond1) + %cond2 = icmp eq i32 %y, -5 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.scmp(i32 %x, i32 %y) + ret i8 %1 +} + +; We can infer ranges based on the location where a UCMP/SCMP result is used +define i8 @scmp_6(i32 noundef %x) { +; CHECK-LABEL: @scmp_6( +; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i32 [[X:%.*]], 10 +; CHECK-NEXT: [[TMP2:%.*]] = select i1 [[TMP1]], i8 -1, i8 5 +; CHECK-NEXT: ret i8 [[TMP2]] +; + %1 = icmp slt i32 %x, 10 + %2 = call i8 @llvm.scmp(i32 %x, i32 10) + %3 = select i1 %1, i8 %2, i8 5 + ret i8 %3 +} + +; Negative test: ranges overlap +define i8 @ucmp_3(i32 %x, i32 %y) { + ; X is within [4, UNSIGNED_MAX) +; CHECK-LABEL: @ucmp_3( +; CHECK-NEXT: [[COND1:%.*]] = icmp uge i32 [[X:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp ult i32 [[Y:%.*]], 6 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.ucmp.i8.i32(i32 [[X]], i32 [[Y]]) +; CHECK-NEXT: ret i8 [[TMP1]] +; + %cond1 = icmp uge i32 %x, 4 + call void @llvm.assume(i1 %cond1) + ; Y is within [0, 6) + %cond2 = icmp ult i32 %y, 6 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.ucmp(i32 %x, i32 %y) + ret i8 %1 +} + +define i8 @scmp_3(i32 %x, i32 %y) { + ; X is within [2, SIGNED_MAX) +; CHECK-LABEL: @scmp_3( +; CHECK-NEXT: [[COND1:%.*]] = icmp sge i32 [[X:%.*]], 2 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp slt i32 [[Y:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.scmp.i8.i32(i32 [[X]], i32 [[Y]]) +; CHECK-NEXT: ret i8 [[TMP1]] +; + %cond1 = icmp sge i32 %x, 2 + call void @llvm.assume(i1 %cond1) + ; Y is within [SIGNED_MIN, 4) + %cond2 = icmp slt i32 %y, 4 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.scmp(i32 %x, i32 %y) + ret i8 %1 +} + +; Negative test: mismatched signedness of range-establishing comparisons and +; of the intrinsic +define i8 @ucmp_4(i32 %x, i32 %y) { + ; X is within [4, SIGNED_MAX) +; CHECK-LABEL: @ucmp_4( +; CHECK-NEXT: [[COND1:%.*]] = icmp sge i32 [[X:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp slt i32 [[Y:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.ucmp.i8.i32(i32 [[X]], i32 [[Y]]) +; CHECK-NEXT: ret i8 [[TMP1]] +; + %cond1 = icmp sge i32 %x, 4 + call void @llvm.assume(i1 %cond1) + ; Y is within [0, 4) + %cond2 = icmp slt i32 %y, 4 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.ucmp(i32 %x, i32 %y) + ret i8 %1 +} + +define i8 @scmp_4(i32 %x, i32 %y) { + ; X is within [4, UNSIGNED_MAX) +; CHECK-LABEL: @scmp_4( +; CHECK-NEXT: [[COND1:%.*]] = icmp uge i32 [[X:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp ult i32 [[Y:%.*]], 4 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.scmp.i8.i32(i32 [[X]], i32 [[Y]]) +; CHECK-NEXT: ret i8 [[TMP1]] +; + %cond1 = icmp uge i32 %x, 4 + call void @llvm.assume(i1 %cond1) + ; Y is within [0, 4) + %cond2 = icmp ult i32 %y, 4 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.scmp(i32 %x, i32 %y) + ret i8 %1 +} + +; Negative test: ranges are the same, but we can't be sure the values are equal +define i8 @ucmp_6(i32 %x, i32 %y) { + ; Both X and Y are within [0, 10] +; CHECK-LABEL: @ucmp_6( +; CHECK-NEXT: [[COND1:%.*]] = icmp ule i32 [[X:%.*]], 10 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND1]]) +; CHECK-NEXT: [[COND2:%.*]] = icmp ule i32 [[Y:%.*]], 10 +; CHECK-NEXT: call void @llvm.assume(i1 [[COND2]]) +; CHECK-NEXT: [[TMP1:%.*]] = call i8 @llvm.ucmp.i8.i32(i32 [[X]], i32 [[Y]]) +; CHECK-NEXT: ret i8 [[TMP1]] +; + %cond1 = icmp ule i32 %x, 10 + call void @llvm.assume(i1 %cond1) + %cond2 = icmp ule i32 %y, 10 + call void @llvm.assume(i1 %cond2) + + %1 = call i8 @llvm.ucmp(i32 %x, i32 %y) + ret i8 %1 +} diff --git a/llvm/test/Transforms/IndVarSimplify/preserving-debugloc-rem-div.ll b/llvm/test/Transforms/IndVarSimplify/preserving-debugloc-rem-div.ll new file mode 100644 index 0000000..0cb8764 --- /dev/null +++ b/llvm/test/Transforms/IndVarSimplify/preserving-debugloc-rem-div.ll @@ -0,0 +1,182 @@ +; Test that the debug information is propagated correctly to the new instructions +; RUN: opt < %s -passes=indvars -S | FileCheck %s + +define void @test_srem_urem(ptr %a) !dbg !5 { +; CHECK-LABEL: define void @test_srem_urem( +; CHECK: [[REM_UREM:%.*]] = urem i32 [[I_01:%.*]], 2, !dbg [[DBG10:![0-9]+]] +; +entry: + br label %for.body, !dbg !8 + +for.body: ; preds = %for.body, %entry + %i.01 = phi i32 [ 0, %entry ], [ %inc, %for.body ], !dbg !9 + %rem = srem i32 %i.01, 2, !dbg !10 + %idxprom = sext i32 %rem to i64, !dbg !11 + %arrayidx = getelementptr inbounds i32, ptr %a, i64 %idxprom, !dbg !12 + store i32 %i.01, ptr %arrayidx, align 4, !dbg !13 + %inc = add nsw i32 %i.01, 1, !dbg !14 + %cmp = icmp slt i32 %inc, 64, !dbg !15 + br i1 %cmp, label %for.body, label %for.end, !dbg !16 + +for.end: ; preds = %for.body + ret void, !dbg !17 +} + +define void @test_sdiv_udiv(ptr %a) !dbg !18 { +; CHECK-LABEL: define void @test_sdiv_udiv( +; CHECK: [[DIV_UDIV:%.*]] = udiv i32 [[I_01:%.*]], 2, !dbg [[DBG21:![0-9]+]] +; +entry: + br label %for.body, !dbg !19 + +for.body: ; preds = %for.body, %entry + %i.01 = phi i32 [ 0, %entry ], [ %inc, %for.body ], !dbg !20 + %div = sdiv i32 %i.01, 2, !dbg !21 + %idxprom = sext i32 %div to i64, !dbg !22 + %arrayidx = getelementptr inbounds i32, ptr %a, i64 %idxprom, !dbg !23 + store i32 %i.01, ptr %arrayidx, align 4, !dbg !24 + %inc = add nsw i32 %i.01, 1, !dbg !25 + %cmp = icmp slt i32 %inc, 64, !dbg !26 + br i1 %cmp, label %for.body, label %for.end, !dbg !27 + +for.end: ; preds = %for.body + ret void, !dbg !28 +} + +; Function Attrs: nounwind +define i32 @test_rem_num_zero(i64 %arg1) #0 !dbg !29 { +; CHECK-LABEL: define i32 @test_rem_num_zero( +; CHECK: [[TMP0:%.*]] = icmp eq i64 [[T12:%.*]], %arg1 +; CHECK: [[IV_REM:%.*]] = select i1 [[TMP0]], i64 0, i64 [[T12]], !dbg [[DBG36:![0-9]+]] +; +bb: + %t = icmp sgt i64 %arg1, 0, !dbg !30 + br i1 %t, label %bb4, label %bb54, !dbg !31 + +bb4: ; preds = %bb + br label %bb5, !dbg !32 + +bb5: ; preds = %bb49, %bb4 + %t6 = phi i64 [ %t51, %bb49 ], [ 0, %bb4 ], !dbg !33 + %t7 = phi i32 [ %t50, %bb49 ], [ 0, %bb4 ], !dbg !34 + %t12 = add nsw i64 %t6, 1, !dbg !35 + %t13 = srem i64 %t12, %arg1, !dbg !36 + %t14 = icmp sgt i64 %arg1, 0, !dbg !37 + br i1 %t14, label %bb15, label %bb49, !dbg !38 + +bb15: ; preds = %bb5 + br label %bb16, !dbg !39 + +bb16: ; preds = %bb44, %bb15 + %t17 = phi i64 [ %t46, %bb44 ], [ 0, %bb15 ], !dbg !40 + %t18 = phi i32 [ %t45, %bb44 ], [ %t7, %bb15 ], !dbg !41 + %t19 = icmp sgt i64 %arg1, 0, !dbg !42 + br i1 %t19, label %bb20, label %bb44, !dbg !43 + +bb20: ; preds = %bb16 + br label %bb21, !dbg !44 + +bb21: ; preds = %bb21, %bb20 + %t25 = mul i64 %t13, %arg1, !dbg !45 + %t42 = icmp slt i64 %t25, %arg1, !dbg !46 + br i1 %t42, label %bb21, label %bb43, !dbg !47 + +bb43: ; preds = %bb21 + br label %bb44, !dbg !48 + +bb44: ; preds = %bb43, %bb16 + %t45 = phi i32 [ %t18, %bb16 ], [ 0, %bb43 ], !dbg !49 + %t46 = add nsw i64 %t17, 1, !dbg !50 + %t47 = icmp slt i64 %t46, %arg1, !dbg !51 + br i1 %t47, label %bb16, label %bb48, !dbg !52 + +bb48: ; preds = %bb44 + br label %bb49, !dbg !53 + +bb49: ; preds = %bb48, %bb5 + %t50 = phi i32 [ %t7, %bb5 ], [ %t45, %bb48 ], !dbg !54 + %t51 = add nsw i64 %t6, 1, !dbg !55 + %t52 = icmp slt i64 %t51, %arg1, !dbg !56 + br i1 %t52, label %bb5, label %bb53, !dbg !57 + +bb53: ; preds = %bb49 + br label %bb54, !dbg !58 + +bb54: ; preds = %bb53, %bb + %t55 = phi i32 [ 0, %bb ], [ %t50, %bb53 ], !dbg !59 + ret i32 %t55, !dbg !60 +} + +attributes #0 = { nounwind } + +!llvm.dbg.cu = !{!0} +!llvm.debugify = !{!2, !3} +!llvm.module.flags = !{!4} + +;. +; CHECK: [[DBG10]] = !DILocation(line: 3, +; CHECK: [[DBG21]] = !DILocation(line: 13, +; CHECK: [[DBG36]] = !DILocation(line: 27, +;. + +!0 = distinct !DICompileUnit(language: DW_LANG_C, file: !1, producer: "debugify", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug) +!1 = !DIFile(filename: "indvars.ll", directory: "/") +!2 = !{i32 51} +!3 = !{i32 0} +!4 = !{i32 2, !"Debug Info Version", i32 3} +!5 = distinct !DISubprogram(name: "test_srem_urem", linkageName: "test_srem_urem", scope: null, file: !1, line: 1, type: !6, scopeLine: 1, spFlags: DISPFlagDefinition | DISPFlagOptimized, unit: !0) +!6 = !DISubroutineType(types: !7) +!7 = !{} +!8 = !DILocation(line: 1, column: 1, scope: !5) +!9 = !DILocation(line: 2, column: 1, scope: !5) +!10 = !DILocation(line: 3, column: 1, scope: !5) +!11 = !DILocation(line: 4, column: 1, scope: !5) +!12 = !DILocation(line: 5, column: 1, scope: !5) +!13 = !DILocation(line: 6, column: 1, scope: !5) +!14 = !DILocation(line: 7, column: 1, scope: !5) +!15 = !DILocation(line: 8, column: 1, scope: !5) +!16 = !DILocation(line: 9, column: 1, scope: !5) +!17 = !DILocation(line: 10, column: 1, scope: !5) +!18 = distinct !DISubprogram(name: "test_sdiv_udiv", linkageName: "test_sdiv_udiv", scope: null, file: !1, line: 11, type: !6, scopeLine: 11, spFlags: DISPFlagDefinition | DISPFlagOptimized, unit: !0) +!19 = !DILocation(line: 11, column: 1, scope: !18) +!20 = !DILocation(line: 12, column: 1, scope: !18) +!21 = !DILocation(line: 13, column: 1, scope: !18) +!22 = !DILocation(line: 14, column: 1, scope: !18) +!23 = !DILocation(line: 15, column: 1, scope: !18) +!24 = !DILocation(line: 16, column: 1, scope: !18) +!25 = !DILocation(line: 17, column: 1, scope: !18) +!26 = !DILocation(line: 18, column: 1, scope: !18) +!27 = !DILocation(line: 19, column: 1, scope: !18) +!28 = !DILocation(line: 20, column: 1, scope: !18) +!29 = distinct !DISubprogram(name: "test_rem_num_zero", linkageName: "test_rem_num_zero", scope: null, file: !1, line: 21, type: !6, scopeLine: 21, spFlags: DISPFlagDefinition | DISPFlagOptimized, unit: !0) +!30 = !DILocation(line: 21, column: 1, scope: !29) +!31 = !DILocation(line: 22, column: 1, scope: !29) +!32 = !DILocation(line: 23, column: 1, scope: !29) +!33 = !DILocation(line: 24, column: 1, scope: !29) +!34 = !DILocation(line: 25, column: 1, scope: !29) +!35 = !DILocation(line: 26, column: 1, scope: !29) +!36 = !DILocation(line: 27, column: 1, scope: !29) +!37 = !DILocation(line: 28, column: 1, scope: !29) +!38 = !DILocation(line: 29, column: 1, scope: !29) +!39 = !DILocation(line: 30, column: 1, scope: !29) +!40 = !DILocation(line: 31, column: 1, scope: !29) +!41 = !DILocation(line: 32, column: 1, scope: !29) +!42 = !DILocation(line: 33, column: 1, scope: !29) +!43 = !DILocation(line: 34, column: 1, scope: !29) +!44 = !DILocation(line: 35, column: 1, scope: !29) +!45 = !DILocation(line: 36, column: 1, scope: !29) +!46 = !DILocation(line: 37, column: 1, scope: !29) +!47 = !DILocation(line: 38, column: 1, scope: !29) +!48 = !DILocation(line: 39, column: 1, scope: !29) +!49 = !DILocation(line: 40, column: 1, scope: !29) +!50 = !DILocation(line: 41, column: 1, scope: !29) +!51 = !DILocation(line: 42, column: 1, scope: !29) +!52 = !DILocation(line: 43, column: 1, scope: !29) +!53 = !DILocation(line: 44, column: 1, scope: !29) +!54 = !DILocation(line: 45, column: 1, scope: !29) +!55 = !DILocation(line: 46, column: 1, scope: !29) +!56 = !DILocation(line: 47, column: 1, scope: !29) +!57 = !DILocation(line: 48, column: 1, scope: !29) +!58 = !DILocation(line: 49, column: 1, scope: !29) +!59 = !DILocation(line: 50, column: 1, scope: !29) +!60 = !DILocation(line: 51, column: 1, scope: !29) diff --git a/llvm/test/Transforms/Inline/always-inline-phase-ordering.ll b/llvm/test/Transforms/Inline/always-inline-phase-ordering.ll index e69ca48..defd1f4 100644 --- a/llvm/test/Transforms/Inline/always-inline-phase-ordering.ll +++ b/llvm/test/Transforms/Inline/always-inline-phase-ordering.ll @@ -6,7 +6,6 @@ target triple = "arm64e-apple-macosx13" ; CHECK: remark: <unknown>:0:0: 'wibble' inlined into 'pluto' with (cost=always): always inline attribute ; CHECK: remark: <unknown>:0:0: 'snork' inlined into 'blam' with (cost=always): always inline attribute ; CHECK: remark: <unknown>:0:0: 'wobble' inlined into 'blam' with (cost=always): always inline attribute -; CHECK: remark: <unknown>:0:0: 'wobble' inlined into 'snork' with (cost=always): always inline attribute ; CHECK: remark: <unknown>:0:0: 'spam' inlined into 'blam' with (cost=65, threshold=75) ; CHECK: remark: <unknown>:0:0: 'wibble.1' inlined into 'widget' with (cost=30, threshold=75) ; CHECK: remark: <unknown>:0:0: 'widget' inlined into 'bar.8' with (cost=30, threshold=75) diff --git a/llvm/test/Transforms/InstCombine/AArch64/sve-intrinsic-comb-no-active-lanes-stores.ll b/llvm/test/Transforms/InstCombine/AArch64/sve-intrinsic-comb-no-active-lanes-stores.ll new file mode 100644 index 0000000..5908dd1 --- /dev/null +++ b/llvm/test/Transforms/InstCombine/AArch64/sve-intrinsic-comb-no-active-lanes-stores.ll @@ -0,0 +1,310 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5 +;RUN: opt -S -passes=instcombine < %s | FileCheck %s +target triple = "aarch64-unknown-linux-gnu" + +define void @test_st1(ptr %a, <vscale x 16 x i8> %b) { +; CHECK-LABEL: define void @test_st1( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 16 x i8> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1.nxv16i8(<vscale x 16 x i8> %b, <vscale x 16 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_st1_scatter(<vscale x 2 x i16> %data_trunc, ptr %base, <vscale x 2 x i64> %b) { +; CHECK-LABEL: define void @test_st1_scatter( +; CHECK-SAME: <vscale x 2 x i16> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1.scatter.nxv2i16(<vscale x 2 x i16> %data_trunc, + <vscale x 2 x i1> zeroinitializer, + ptr %base, + <vscale x 2 x i64> %b) + ret void +} + +define void @test_st1_scatter_index(<vscale x 2 x i32> %data_trunc, ptr %base, <vscale x 2 x i64> %offsets) { +; CHECK-LABEL: define void @test_st1_scatter_index( +; CHECK-SAME: <vscale x 2 x i32> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: call void @llvm.aarch64.sve.st1.scatter.index.nxv2i32(<vscale x 2 x i32> [[DATA_TRUNC]], <vscale x 2 x i1> zeroinitializer, ptr [[BASE]], <vscale x 2 x i64> [[OFFSETS]]) +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1.scatter.index.nxv2i32(<vscale x 2 x i32> %data_trunc, + <vscale x 2 x i1> zeroinitializer, + ptr %base, + <vscale x 2 x i64> %offsets) + ret void +} + +define void @test_st1_scatter_scalar_offset(<vscale x 4 x i8> %data_trunc, <vscale x 4 x i32> %base) { +; CHECK-LABEL: define void @test_st1_scatter_scalar_offset( +; CHECK-SAME: <vscale x 4 x i8> [[DATA_TRUNC:%.*]], <vscale x 4 x i32> [[BASE:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv4i8.nxv4i32(<vscale x 4 x i8> %data_trunc, + <vscale x 4 x i1> zeroinitializer, + <vscale x 4 x i32> %base, + i64 16) + ret void +} + +define void @test_st1_scatter_sxtw(<vscale x 4 x i8> %data_trunc, ptr %base, <vscale x 4 x i32> %offsets) { +; CHECK-LABEL: define void @test_st1_scatter_sxtw( +; CHECK-SAME: <vscale x 4 x i8> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]]) { +; CHECK-NEXT: ret void +; + call void @llvm.aarch64.sve.st1.scatter.sxtw.nxv4i8(<vscale x 4 x i8> %data_trunc, + <vscale x 4 x i1> zeroinitializer, + ptr %base, + <vscale x 4 x i32> %offsets) + ret void +} + +define void @test_st1_scatter_sxtw_index(<vscale x 4 x i16> %data_trunc, ptr %base, <vscale x 4 x i32> %indices) { +; CHECK-LABEL: define void @test_st1_scatter_sxtw_index( +; CHECK-SAME: <vscale x 4 x i16> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]]) { +; CHECK-NEXT: ret void +; + call void @llvm.aarch64.sve.st1.scatter.sxtw.index.nxv4i16(<vscale x 4 x i16> %data_trunc, + <vscale x 4 x i1> zeroinitializer, + ptr %base, + <vscale x 4 x i32> %indices) + ret void +} + +define void @test_st1_scatter_uxtw(<vscale x 4 x i8> %data_trunc, ptr %base, <vscale x 4 x i32> %offsets) { +; CHECK-LABEL: define void @test_st1_scatter_uxtw( +; CHECK-SAME: <vscale x 4 x i8> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]]) { +; CHECK-NEXT: ret void +; + call void @llvm.aarch64.sve.st1.scatter.uxtw.nxv4i8(<vscale x 4 x i8> %data_trunc, + <vscale x 4 x i1> zeroinitializer, + ptr %base, + <vscale x 4 x i32> %offsets) + ret void +} + +define void @test_st1_scatter_uxtw_index(<vscale x 4 x i16> %data_trunc, ptr %base, <vscale x 4 x i32> %indices) { +; CHECK-LABEL: define void @test_st1_scatter_uxtw_index( +; CHECK-SAME: <vscale x 4 x i16> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]]) { +; CHECK-NEXT: ret void +; + call void @llvm.aarch64.sve.st1.scatter.uxtw.index.nxv4i16(<vscale x 4 x i16> %data_trunc, + <vscale x 4 x i1> zeroinitializer, + ptr %base, + <vscale x 4 x i32> %indices) + ret void +} + +define void @test_st1dq(<vscale x 2 x i64> %zt, ptr %gep1) { +; CHECK-LABEL: define void @test_st1dq( +; CHECK-SAME: <vscale x 2 x i64> [[ZT:%.*]], ptr [[GEP1:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1dq.nxv2i64(<vscale x 2 x i64> %zt, <vscale x 1 x i1> zeroinitializer, ptr %gep1) + ret void +} + +define void @test_st1q_scatter_index(<vscale x 8 x i16> %data, <vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %idx) { +; CHECK-LABEL: define void @test_st1q_scatter_index( +; CHECK-SAME: <vscale x 8 x i16> [[DATA:%.*]], <vscale x 1 x i1> [[PG:%.*]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[IDX:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1q.scatter.index.nxv8i16(<vscale x 8 x i16> %data, <vscale x 1 x i1> zeroinitializer, ptr %base, <vscale x 2 x i64> %idx) + ret void +} + +define void @test_st1q_scatter_scalar_offset(<vscale x 2 x i64> %data, <vscale x 2 x i64> %base) { +; CHECK-LABEL: define void @test_st1q_scatter_scalar_offset( +; CHECK-SAME: <vscale x 2 x i64> [[DATA:%.*]], <vscale x 2 x i64> [[BASE:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1q.scatter.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i64> %data, <vscale x 1 x i1> zeroinitializer, <vscale x 2 x i64> %base, i64 0) + ret void +} + +define void @test_st1q_scatter_vector_offset(<vscale x 8 x i16> %data, ptr %base, <vscale x 2 x i64> %off) { +; CHECK-LABEL: define void @test_st1q_scatter_vector_offset( +; CHECK-SAME: <vscale x 8 x i16> [[DATA:%.*]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[OFF:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1q.scatter.vector.offset.nxv8i16(<vscale x 8 x i16> %data, <vscale x 1 x i1> zeroinitializer, ptr %base, <vscale x 2 x i64> %off) + ret void +} + +define void @test_st1wq(ptr %a, <vscale x 4 x i32> %b) { +; CHECK-LABEL: define void @test_st1wq( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 4 x i32> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.st1wq.nxv4i32(<vscale x 4 x i32> %b, <vscale x 1 x i1> zeroinitializer, ptr %a) + ret void +} + + +define void @test_st2(ptr %a, <vscale x 8 x i32> %b) { +; CHECK-LABEL: define void @test_st2( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 8 x i32> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + %0 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> %b, i64 0) + %1 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> %b, i64 4) + tail call void @llvm.aarch64.sve.st2.nxv4i32(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_st2q(ptr %a, <vscale x 8 x i32> %b) { +; CHECK-LABEL: define void @test_st2q( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 8 x i32> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + %0 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> %b, i64 0) + %1 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> %b, i64 4) + tail call void @llvm.aarch64.sve.st2q.nxv4i32(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_st3(ptr %a, <vscale x 12 x i32> %b) { +; CHECK-LABEL: define void @test_st3( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 12 x i32> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + %0 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv12i32(<vscale x 12 x i32> %b, i64 0) + %1 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv12i32(<vscale x 12 x i32> %b, i64 4) + %2 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv12i32(<vscale x 12 x i32> %b, i64 8) + tail call void @llvm.aarch64.sve.st3.nxv4i32(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_st3q(ptr %a, <vscale x 12 x i32> %b) { +; CHECK-LABEL: define void @test_st3q( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 12 x i32> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + %0 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv12i32(<vscale x 12 x i32> %b, i64 0) + %1 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv12i32(<vscale x 12 x i32> %b, i64 4) + %2 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv12i32(<vscale x 12 x i32> %b, i64 8) + tail call void @llvm.aarch64.sve.st3q.nxv4i32(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_st4(ptr %a, <vscale x 16 x i32> %b) { +; CHECK-LABEL: define void @test_st4( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 16 x i32> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + %0 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 0) + %1 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 4) + %2 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 8) + %3 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 12) + tail call void @llvm.aarch64.sve.st4.nxv4i32(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, <vscale x 4 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_st4q(ptr %a, <vscale x 16 x i32> %b) { +; CHECK-LABEL: define void @test_st4q( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 16 x i32> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + %0 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 0) + %1 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 4) + %2 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 8) + %3 = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %b, i64 12) + tail call void @llvm.aarch64.sve.st4q.nxv4i32(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, <vscale x 4 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_stnt1(ptr %a, <vscale x 16 x i8> %b) { +; CHECK-LABEL: define void @test_stnt1( +; CHECK-SAME: ptr [[A:%.*]], <vscale x 16 x i8> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.stnt1.nxv16i8(<vscale x 16 x i8> %b, <vscale x 16 x i1> zeroinitializer, ptr %a) + ret void +} + +define void @test_stnt1_scatter(<vscale x 2 x i16> %data_trunc, ptr %base, <vscale x 2 x i64> %b) { +; CHECK-LABEL: define void @test_stnt1_scatter( +; CHECK-SAME: <vscale x 2 x i16> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[B:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.stnt1.scatter.nxv2i16(<vscale x 2 x i16> %data_trunc, + <vscale x 2 x i1> zeroinitializer, + ptr %base, + <vscale x 2 x i64> %b) + ret void +} + +define void @test_stnt1_scatter_index(<vscale x 2 x i32> %data_trunc, ptr %base, <vscale x 2 x i64> %offsets) { +; CHECK-LABEL: define void @test_stnt1_scatter_index( +; CHECK-SAME: <vscale x 2 x i32> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.stnt1.scatter.index.nxv2i32(<vscale x 2 x i32> %data_trunc, + <vscale x 2 x i1> zeroinitializer, + ptr %base, + <vscale x 2 x i64> %offsets) + ret void +} + +define void @test_stnt1_scatter_scalar_offset(<vscale x 4 x i8> %data_trunc, <vscale x 4 x i32> %base) { +; CHECK-LABEL: define void @test_stnt1_scatter_scalar_offset( +; CHECK-SAME: <vscale x 4 x i8> [[DATA_TRUNC:%.*]], <vscale x 4 x i32> [[BASE:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: ret void +; +entry: + call void @llvm.aarch64.sve.stnt1.scatter.scalar.offset.nxv4i8.nxv4i32(<vscale x 4 x i8> %data_trunc, + <vscale x 4 x i1> zeroinitializer, + <vscale x 4 x i32> %base, + i64 16) + ret void +} + +define void @test_stnt1_scatter_uxtw(<vscale x 4 x i8> %data_trunc, ptr %base, <vscale x 4 x i32> %offsets) { +; CHECK-LABEL: define void @test_stnt1_scatter_uxtw( +; CHECK-SAME: <vscale x 4 x i8> [[DATA_TRUNC:%.*]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]]) { +; CHECK-NEXT: ret void +; + call void @llvm.aarch64.sve.stnt1.scatter.uxtw.nxv4i8(<vscale x 4 x i8> %data_trunc, + <vscale x 4 x i1> zeroinitializer, + ptr %base, + <vscale x 4 x i32> %offsets) + ret void +} diff --git a/llvm/test/Transforms/InstCombine/known-bits.ll b/llvm/test/Transforms/InstCombine/known-bits.ll index d7a8386..c7445a6 100644 --- a/llvm/test/Transforms/InstCombine/known-bits.ll +++ b/llvm/test/Transforms/InstCombine/known-bits.ll @@ -2025,6 +2025,34 @@ define i8 @simplifydemanded_context(i8 %x, i8 %y) { ret i8 %and2 } +define i16 @pr97330(i1 %c, ptr %p1, ptr %p2) { +; CHECK-LABEL: @pr97330( +; CHECK-NEXT: entry: +; CHECK-NEXT: br i1 [[C:%.*]], label [[EXIT:%.*]], label [[IF:%.*]] +; CHECK: if: +; CHECK-NEXT: unreachable +; CHECK: exit: +; CHECK-NEXT: [[V:%.*]] = load i64, ptr [[P1:%.*]], align 8 +; CHECK-NEXT: [[CONV:%.*]] = trunc i64 [[V]] to i16 +; CHECK-NEXT: ret i16 [[CONV]] +; +entry: + %v = load i64, ptr %p1, align 8 + %conv = trunc i64 %v to i16 + br i1 %c, label %exit, label %if + +if: + %cmp = icmp ne i16 %conv, 1 + %conv2 = zext i1 %cmp to i32 + store i32 %conv2, ptr %p2, align 4 + %cmp2 = icmp eq i64 %v, 1 + call void @llvm.assume(i1 %cmp2) + unreachable + +exit: + ret i16 %conv +} + declare void @dummy() declare void @use(i1) declare void @sink(i8) diff --git a/llvm/test/Transforms/InstCombine/ldexp.ll b/llvm/test/Transforms/InstCombine/ldexp.ll index 3ede7d9..7ae483c 100644 --- a/llvm/test/Transforms/InstCombine/ldexp.ll +++ b/llvm/test/Transforms/InstCombine/ldexp.ll @@ -870,7 +870,7 @@ define float @ldexp_2_flags(float %x) { define float @ldexp_metadata(float %x) { ; CHECK-LABEL: define float @ldexp_metadata ; CHECK-SAME: (float [[X:%.*]]) { -; CHECK-NEXT: [[LDEXP:%.*]] = call float @llvm.ldexp.f32.i32(float [[X]], i32 2), !foo !2 +; CHECK-NEXT: [[LDEXP:%.*]] = call float @llvm.ldexp.f32.i32(float [[X]], i32 2), !foo [[META2:![0-9]+]] ; CHECK-NEXT: ret float [[LDEXP]] ; %ldexp = call float @llvm.ldexp.f32.i32(float %x, i32 2), !foo !2 @@ -889,6 +889,132 @@ define float @ldexp_8_contractable(float %x, float %y) { ret float %fadd } +define float @ldexp_f32_mask_select_0(i1 %cond, float %x, i32 %y) { +; CHECK-LABEL: define float @ldexp_f32_mask_select_0 +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]]) { +; CHECK-NEXT: [[TMP1:%.*]] = call float @llvm.ldexp.f32.i32(float [[X]], i32 [[Y]]) +; CHECK-NEXT: [[LDEXP:%.*]] = select i1 [[COND]], float [[TMP1]], float [[X]] +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 %y, i32 0 + %ldexp = call float @llvm.ldexp.f32.i32(float %x, i32 %select) + ret float %ldexp +} + +define float @ldexp_nnan_f32_mask_select_0(i1 %cond, float %x, i32 %y) { +; CHECK-LABEL: define float @ldexp_nnan_f32_mask_select_0 +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]]) { +; CHECK-NEXT: [[TMP1:%.*]] = call nnan float @llvm.ldexp.f32.i32(float [[X]], i32 [[Y]]) +; CHECK-NEXT: [[LDEXP:%.*]] = select i1 [[COND]], float [[TMP1]], float [[X]] +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 %y, i32 0 + %ldexp = call nnan float @llvm.ldexp.f32.i32(float %x, i32 %select) + ret float %ldexp +} + +define float @ldexp_flags_f32_mask_select_0(i1 %cond, float %x, i32 %y) { +; CHECK-LABEL: define float @ldexp_flags_f32_mask_select_0 +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]]) { +; CHECK-NEXT: [[TMP1:%.*]] = call ninf nsz float @llvm.ldexp.f32.i32(float [[X]], i32 [[Y]]) +; CHECK-NEXT: [[LDEXP:%.*]] = select i1 [[COND]], float [[TMP1]], float [[X]] +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 %y, i32 0 + %ldexp = call nsz ninf float @llvm.ldexp.f32.i32(float %x, i32 %select) + ret float %ldexp +} + +define float @ldexp_f32_mask_select_0_swap(i1 %cond, float %x, i32 %y) { +; CHECK-LABEL: define float @ldexp_f32_mask_select_0_swap +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]]) { +; CHECK-NEXT: [[TMP1:%.*]] = call float @llvm.ldexp.f32.i32(float [[X]], i32 [[Y]]) +; CHECK-NEXT: [[LDEXP:%.*]] = select i1 [[COND]], float [[X]], float [[TMP1]] +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 0, i32 %y + %ldexp = call float @llvm.ldexp.f32.i32(float %x, i32 %select) + ret float %ldexp +} + +define float @ldexp_f32_mask_select_1(i1 %cond, float %x, i32 %y) { +; CHECK-LABEL: define float @ldexp_f32_mask_select_1 +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]]) { +; CHECK-NEXT: [[SELECT:%.*]] = select i1 [[COND]], i32 [[Y]], i32 1 +; CHECK-NEXT: [[LDEXP:%.*]] = call float @llvm.ldexp.f32.i32(float [[X]], i32 [[SELECT]]) +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 %y, i32 1 + %ldexp = call float @llvm.ldexp.f32.i32(float %x, i32 %select) + ret float %ldexp +} + +define float @ldexp_f32_mask_select_0_multi_use(i1 %cond, float %x, i32 %y, ptr %ptr) { +; CHECK-LABEL: define float @ldexp_f32_mask_select_0_multi_use +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]], ptr [[PTR:%.*]]) { +; CHECK-NEXT: [[SELECT:%.*]] = select i1 [[COND]], i32 [[Y]], i32 0 +; CHECK-NEXT: store i32 [[SELECT]], ptr [[PTR]], align 4 +; CHECK-NEXT: [[LDEXP:%.*]] = call float @llvm.ldexp.f32.i32(float [[X]], i32 [[SELECT]]) +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 %y, i32 0 + store i32 %select, ptr %ptr + %ldexp = call float @llvm.ldexp.f32.i32(float %x, i32 %select) + ret float %ldexp +} + +define float @ldexp_f32_mask_select_0_swap_multi_use(i1 %cond, float %x, i32 %y, ptr %ptr) { +; CHECK-LABEL: define float @ldexp_f32_mask_select_0_swap_multi_use +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]], ptr [[PTR:%.*]]) { +; CHECK-NEXT: [[SELECT:%.*]] = select i1 [[COND]], i32 0, i32 [[Y]] +; CHECK-NEXT: store i32 [[SELECT]], ptr [[PTR]], align 4 +; CHECK-NEXT: [[LDEXP:%.*]] = call float @llvm.ldexp.f32.i32(float [[X]], i32 [[SELECT]]) +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 0, i32 %y + store i32 %select, ptr %ptr + %ldexp = call float @llvm.ldexp.f32.i32(float %x, i32 %select) + ret float %ldexp +} + +define float @ldexp_f32_mask_select_0_strictfp(i1 %cond, float %x, i32 %y) #0 { +; CHECK-LABEL: define float @ldexp_f32_mask_select_0_strictfp +; CHECK-SAME: (i1 [[COND:%.*]], float [[X:%.*]], i32 [[Y:%.*]]) #[[ATTR1:[0-9]+]] { +; CHECK-NEXT: [[SELECT:%.*]] = select i1 [[COND]], i32 [[Y]], i32 0 +; CHECK-NEXT: [[LDEXP:%.*]] = call float @llvm.experimental.constrained.ldexp.f32.i32(float [[X]], i32 [[SELECT]], metadata !"round.dynamic", metadata !"fpexcept.strict") +; CHECK-NEXT: ret float [[LDEXP]] +; + %select = select i1 %cond, i32 %y, i32 0 + %ldexp = call float @llvm.experimental.constrained.ldexp.f32.i32(float %x, i32 %select, metadata !"round.dynamic", metadata !"fpexcept.strict") + ret float %ldexp +} + +define <2 x float> @ldexp_v2f32_mask_select_0(<2 x i1> %cond, <2 x float> %x, <2 x i32> %y) { +; CHECK-LABEL: define <2 x float> @ldexp_v2f32_mask_select_0 +; CHECK-SAME: (<2 x i1> [[COND:%.*]], <2 x float> [[X:%.*]], <2 x i32> [[Y:%.*]]) { +; CHECK-NEXT: [[TMP1:%.*]] = call nnan nsz <2 x float> @llvm.ldexp.v2f32.v2i32(<2 x float> [[X]], <2 x i32> [[Y]]) +; CHECK-NEXT: [[LDEXP:%.*]] = select <2 x i1> [[COND]], <2 x float> [[TMP1]], <2 x float> [[X]] +; CHECK-NEXT: ret <2 x float> [[LDEXP]] +; + %select = select <2 x i1> %cond, <2 x i32> %y, <2 x i32> zeroinitializer + %ldexp = call nsz nnan <2 x float> @llvm.ldexp.f32.v2i32(<2 x float> %x, <2 x i32> %select) + ret <2 x float> %ldexp +} + +define <2 x float> @ldexp_v2f32_mask_select_0_swap(<2 x i1> %cond, <2 x float> %x, <2 x i32> %y) { +; CHECK-LABEL: define <2 x float> @ldexp_v2f32_mask_select_0_swap +; CHECK-SAME: (<2 x i1> [[COND:%.*]], <2 x float> [[X:%.*]], <2 x i32> [[Y:%.*]]) { +; CHECK-NEXT: [[TMP1:%.*]] = call nnan nsz <2 x float> @llvm.ldexp.v2f32.v2i32(<2 x float> [[X]], <2 x i32> [[Y]]) +; CHECK-NEXT: [[LDEXP:%.*]] = select <2 x i1> [[COND]], <2 x float> [[X]], <2 x float> [[TMP1]] +; CHECK-NEXT: ret <2 x float> [[LDEXP]] +; + %select = select <2 x i1> %cond, <2 x i32> zeroinitializer, <2 x i32> %y + %ldexp = call nsz nnan <2 x float> @llvm.ldexp.f32.v2i32(<2 x float> %x, <2 x i32> %select) + ret <2 x float> %ldexp +} + +attributes #0 = { strictfp } + !0 = !{i32 -127, i32 0} !1 = !{i32 0, i32 127} !2 = !{} diff --git a/llvm/test/Transforms/PhaseOrdering/generate-fabs.ll b/llvm/test/Transforms/PhaseOrdering/generate-fabs.ll new file mode 100644 index 0000000..25ac510 --- /dev/null +++ b/llvm/test/Transforms/PhaseOrdering/generate-fabs.ll @@ -0,0 +1,29 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5 +; RUN: opt -passes='default<O1>' -S < %s | FileCheck %s + +; Following test must generate fabs intrinsic. It goes through following stages +; 1. SROA propagates the nsz function attribute on the phi node. +; 2. SimplifyCFG pass converts phi node to select. +; 3. InstCombine converts select with nsz flag into fabs intrinsic. + +define double @fabs_fcmp_olt_nsz_func_attr(double %0, double %1) "no-signed-zeros-fp-math"="true" { +; CHECK-LABEL: define double @fabs_fcmp_olt_nsz_func_attr( +; CHECK-SAME: double [[TMP0:%.*]], double [[TMP1:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[X_0:%.*]] = tail call nnan nsz double @llvm.fabs.f64(double [[TMP0]]) +; CHECK-NEXT: ret double [[X_0]] +entry: + %x = alloca double + store double %0, ptr %x + %cmp = fcmp nnan nsz olt double %0, 0.000000e+00 + br i1 %cmp, label %if.then, label %return + +if.then: ; preds = %entry + %fneg = fneg nnan nsz double %0 + store double %fneg, ptr %x + br label %return + +return: ; preds = %entry, %if.then + %ret = load double, ptr %x + ret double %ret +} diff --git a/llvm/test/Transforms/SLPVectorizer/AArch64/uselistorder.ll b/llvm/test/Transforms/SLPVectorizer/AArch64/uselistorder.ll new file mode 100644 index 0000000..3a68a37 --- /dev/null +++ b/llvm/test/Transforms/SLPVectorizer/AArch64/uselistorder.ll @@ -0,0 +1,43 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; RUN: opt < %s -passes=slp-vectorizer -S -pass-remarks-missed=slp-vectorizer 2>&1 | FileCheck %s + +target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128" +target triple = "aarch64-unknown-linux-gnu" + +; This test has UB but the crash in #95016 only happens with it +define void @uselistorder_test() { +; CHECK-LABEL: @uselistorder_test( +; CHECK-NEXT: [[TMP1:%.*]] = insertelement <2 x double> poison, double 0.000000e+00, i32 0 +; CHECK-NEXT: [[TMP2:%.*]] = insertelement <2 x double> [[TMP1]], double 0.000000e+00, i32 1 +; CHECK-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[TMP2]], zeroinitializer +; CHECK-NEXT: [[TMP4:%.*]] = fmul <2 x double> zeroinitializer, [[TMP3]] +; CHECK-NEXT: [[TMP5:%.*]] = fmul <2 x double> [[TMP4]], zeroinitializer +; CHECK-NEXT: [[TMP6:%.*]] = select <2 x i1> zeroinitializer, <2 x double> zeroinitializer, <2 x double> [[TMP5]] +; CHECK-NEXT: [[TMP7:%.*]] = fmul <2 x double> [[TMP6]], zeroinitializer +; CHECK-NEXT: [[TMP8:%.*]] = fadd <2 x double> [[TMP7]], zeroinitializer +; CHECK-NEXT: store <2 x double> [[TMP8]], ptr null, align 8 +; CHECK-NEXT: ret void +; + %max1 = call double @llvm.maximum.f64(double 0.000000e+00, double 0.000000e+00) [ "a_list"(ptr null) ] + %add1 = fadd double %max1, 0.000000e+00 + %mul1 = fmul double 0.000000e+00, %add1 + %mul2 = fmul double %mul1, 0.000000e+00 + %sel1 = select i1 false, double 0.000000e+00, double %mul2 + %max2 = call double @llvm.maximum.f64(double 0.000000e+00, double 0.000000e+00) + %add2 = fadd double %max2, 0.000000e+00 + %mul3 = fmul double 0.000000e+00, %add2 + %mul4 = fmul double %mul3, 0.000000e+00 + %sel2 = select i1 false, double 0.000000e+00, double %mul4 + %mul5 = fmul double %sel2, 0.000000e+00 + %add3 = fadd double 0.000000e+00, %mul5 + %gep1 = getelementptr { double, [1 x [2 x double]] }, ptr null, i64 0, i32 1 + store double %add3, ptr %gep1, align 8 + %mul6 = fmul double %sel1, 0.000000e+00 + %add4 = fadd double %mul6, 0.000000e+00 + store double %add4, ptr null, align 8 + ret void +} + +declare double @llvm.maximum.f64(double, double) #0 + +attributes #0 = { nocallback nofree nosync nounwind speculatable willreturn memory(none) } diff --git a/llvm/test/Transforms/SROA/propagate-fast-math-flags-on-phi.ll b/llvm/test/Transforms/SROA/propagate-fast-math-flags-on-phi.ll new file mode 100644 index 0000000..2cc2636 --- /dev/null +++ b/llvm/test/Transforms/SROA/propagate-fast-math-flags-on-phi.ll @@ -0,0 +1,79 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4 +; RUN: opt < %s -passes='sroa' -S | FileCheck %s +define double @phi_with_nsz(double %x) "no-signed-zeros-fp-math"="true" { +; CHECK-LABEL: define double @phi_with_nsz( +; CHECK-SAME: double [[X:%.*]]) #[[ATTR0:[0-9]+]] { +; CHECK-NEXT: entry: +; CHECK-NEXT: [[CMP:%.*]] = fcmp olt double [[X]], 0.000000e+00 +; CHECK-NEXT: br i1 [[CMP]], label [[IF_THEN:%.*]], label [[RETURN:%.*]] +; CHECK: if.then: +; CHECK-NEXT: [[FNEG:%.*]] = fneg double [[X]] +; CHECK-NEXT: br label [[RETURN]] +; CHECK: return: +; CHECK-NEXT: [[X_ADDR_0:%.*]] = phi nsz double [ [[FNEG]], [[IF_THEN]] ], [ undef, [[ENTRY:%.*]] ] +; CHECK-NEXT: ret double [[X_ADDR_0]] +entry: + %x.addr = alloca double + %cmp = fcmp olt double %x, 0.0 + br i1 %cmp, label %if.then, label %return + +if.then: ; preds = %entry + %fneg = fneg double %x + store double %fneg, ptr %x.addr + br label %return + +return: ; preds = %entry,%if.then + %retval = load double, ptr %x.addr + ret double %retval +} + +define <2 x double> @vector_phi_with_nsz(<2 x double> %x, i1 %cmp, <2 x double> %a, <2 x double> %b) "no-signed-zeros-fp-math"="true" { +; CHECK-LABEL: define <2 x double> @vector_phi_with_nsz( +; CHECK-SAME: <2 x double> [[X:%.*]], i1 [[CMP:%.*]], <2 x double> [[A:%.*]], <2 x double> [[B:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: entry: +; CHECK-NEXT: br i1 [[CMP]], label [[IF_THEN:%.*]], label [[RETURN:%.*]] +; CHECK: if.then: +; CHECK-NEXT: br label [[RETURN]] +; CHECK: return: +; CHECK-NEXT: [[X_ADDR_0:%.*]] = phi nsz <2 x double> [ [[B]], [[IF_THEN]] ], [ [[A]], [[ENTRY:%.*]] ] +; CHECK-NEXT: ret <2 x double> [[X_ADDR_0]] +entry: + %x.addr = alloca <2 x double> + store <2 x double> %a, ptr %x.addr + br i1 %cmp, label %if.then, label %return + +if.then: ; preds = %entry + store <2 x double> %b, ptr %x.addr + br label %return + +return: ; preds = %entry,%if.then + %retval = load <2 x double>, ptr %x.addr + ret <2 x double> %retval +} + +define double @phi_without_nsz(double %x) "no-signed-zeros-fp-math"="false" { +; CHECK-LABEL: define double @phi_without_nsz( +; CHECK-SAME: double [[X:%.*]]) #[[ATTR1:[0-9]+]] { +; CHECK-NEXT: entry: +; CHECK-NEXT: [[CMP:%.*]] = fcmp olt double [[X]], 0.000000e+00 +; CHECK-NEXT: br i1 [[CMP]], label [[IF_THEN:%.*]], label [[RETURN:%.*]] +; CHECK: if.then: +; CHECK-NEXT: [[FNEG:%.*]] = fneg double [[X]] +; CHECK-NEXT: br label [[RETURN]] +; CHECK: return: +; CHECK-NEXT: [[X_ADDR_0:%.*]] = phi double [ [[FNEG]], [[IF_THEN]] ], [ undef, [[ENTRY:%.*]] ] +; CHECK-NEXT: ret double [[X_ADDR_0]] +entry: + %x.addr = alloca double + %cmp = fcmp olt double %x, 0.0 + br i1 %cmp, label %if.then, label %return + +if.then: ; preds = %entry + %fneg = fneg double %x + store double %fneg, ptr %x.addr + br label %return + +return: ; preds = %entry,%if.then + %retval = load double, ptr %x.addr + ret double %retval +} diff --git a/llvm/tools/dsymutil/MachOUtils.cpp b/llvm/tools/dsymutil/MachOUtils.cpp index fba6630..d2bdcf8 100644 --- a/llvm/tools/dsymutil/MachOUtils.cpp +++ b/llvm/tools/dsymutil/MachOUtils.cpp @@ -12,7 +12,6 @@ #include "LinkUtils.h" #include "llvm/ADT/SmallString.h" #include "llvm/CodeGen/NonRelocatableStringpool.h" -#include "llvm/MC/MCAsmLayout.h" #include "llvm/MC/MCAssembler.h" #include "llvm/MC/MCMachObjectWriter.h" #include "llvm/MC/MCObjectStreamer.h" @@ -381,8 +380,7 @@ bool generateDsymCompanion( auto &Writer = static_cast<MachObjectWriter &>(MCAsm.getWriter()); // Layout but don't emit. - MCAsmLayout Layout(MCAsm); - MCAsm.layout(Layout); + MCAsm.layout(); BinaryHolder InputBinaryHolder(VFS, false); diff --git a/llvm/utils/gn/secondary/lldb/source/Commands/BUILD.gn b/llvm/utils/gn/secondary/lldb/source/Commands/BUILD.gn index f19f948..164ee90 100644 --- a/llvm/utils/gn/secondary/lldb/source/Commands/BUILD.gn +++ b/llvm/utils/gn/secondary/lldb/source/Commands/BUILD.gn @@ -52,7 +52,7 @@ static_library("Commands") { "CommandObjectQuit.cpp", "CommandObjectRegexCommand.cpp", "CommandObjectRegister.cpp", - "CommandObjectScript.cpp", + "CommandObjectScripting.cpp", "CommandObjectSession.cpp", "CommandObjectSettings.cpp", "CommandObjectSource.cpp", diff --git a/llvm/utils/gn/secondary/llvm/lib/SandboxIR/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/SandboxIR/BUILD.gn new file mode 100644 index 0000000..4650741 --- /dev/null +++ b/llvm/utils/gn/secondary/llvm/lib/SandboxIR/BUILD.gn @@ -0,0 +1,8 @@ +static_library("SandboxIR") { + output_name = "LLVMSandboxIR" + deps = [ + "//llvm/lib/IR", + "//llvm/lib/Support", + ] + sources = [ "SandboxIR.cpp" ] +} diff --git a/llvm/utils/gn/secondary/llvm/unittests/BUILD.gn b/llvm/utils/gn/secondary/llvm/unittests/BUILD.gn index 30e49ff..92c6329 100644 --- a/llvm/utils/gn/secondary/llvm/unittests/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/unittests/BUILD.gn @@ -43,6 +43,7 @@ group("unittests") { "Passes/Plugins:PluginsTests", "ProfileData:ProfileDataTests", "Remarks:RemarksTests", + "SandboxIR:SandboxIRTests", "Support:SupportTests", "Support/DynamicLibrary:DynamicLibraryTests", "TableGen:TableGenTests", diff --git a/llvm/utils/gn/secondary/llvm/unittests/SandboxIR/BUILD.gn b/llvm/utils/gn/secondary/llvm/unittests/SandboxIR/BUILD.gn new file mode 100644 index 0000000..d7164f2 --- /dev/null +++ b/llvm/utils/gn/secondary/llvm/unittests/SandboxIR/BUILD.gn @@ -0,0 +1,10 @@ +import("//third-party/unittest/unittest.gni") + +unittest("SandboxIRTests") { + deps = [ + "//llvm/lib/AsmParser", + "//llvm/lib/IR", + "//llvm/lib/SandboxIR", + ] + sources = [ "SandboxIRTest.cpp" ] +} diff --git a/mlir/cmake/modules/AddMLIRPython.cmake b/mlir/cmake/modules/AddMLIRPython.cmake index f030798..7b91f43 100644 --- a/mlir/cmake/modules/AddMLIRPython.cmake +++ b/mlir/cmake/modules/AddMLIRPython.cmake @@ -565,8 +565,6 @@ function(add_mlir_python_sources_target name) message(FATAL_ERROR "Unhandled arguments to add_mlir_python_sources_target(${name}, ... : ${ARG_UNPARSED_ARGUMENTS}") endif() - add_custom_target(${name}) - # On Windows create_symlink requires special permissions. Use copy_if_different instead. if(CMAKE_HOST_WIN32) set(_link_or_copy copy_if_different) @@ -575,8 +573,6 @@ function(add_mlir_python_sources_target name) endif() foreach(_sources_target ${ARG_SOURCES_TARGETS}) - add_dependencies(${name} ${_sources_target}) - get_target_property(_src_paths ${_sources_target} SOURCES) if(NOT _src_paths) get_target_property(_src_paths ${_sources_target} INTERFACE_SOURCES) @@ -590,6 +586,8 @@ function(add_mlir_python_sources_target name) get_target_property(_root_dir ${_sources_target} INTERFACE_INCLUDE_DIRECTORIES) endif() + # Initialize an empty list of all Python source destination paths. + set(all_dest_paths "") foreach(_src_path ${_src_paths}) file(RELATIVE_PATH _source_relative_path "${_root_dir}" "${_src_path}") set(_dest_path "${ARG_OUTPUT_DIRECTORY}/${_source_relative_path}") @@ -598,13 +596,17 @@ function(add_mlir_python_sources_target name) file(MAKE_DIRECTORY "${_dest_dir}") add_custom_command( - TARGET ${name} PRE_BUILD + OUTPUT "${_dest_path}" + PRE_BUILD COMMENT "Copying python source ${_src_path} -> ${_dest_path}" DEPENDS "${_src_path}" - BYPRODUCTS "${_dest_path}" COMMAND "${CMAKE_COMMAND}" -E ${_link_or_copy} "${_src_path}" "${_dest_path}" ) + + # Track the symlink or copy command output. + list(APPEND all_dest_paths "${_dest_path}") + if(ARG_INSTALL_DIR) # We have to install each file individually because we need to preserve # the relative directory structure in the install destination. @@ -621,6 +623,9 @@ function(add_mlir_python_sources_target name) endif() endforeach() endforeach() + + # Create a new custom target that depends on all symlinked or copied sources. + add_custom_target("${name}" DEPENDS ${all_dest_paths}) endfunction() ################################################################################ diff --git a/mlir/docs/PDLL.md b/mlir/docs/PDLL.md index 340940f..f565115 100644 --- a/mlir/docs/PDLL.md +++ b/mlir/docs/PDLL.md @@ -277,7 +277,7 @@ Pattern { // * Match Section // - Describes the input IR. let root = op<toy.reshape>(op<toy.reshape>(arg: Value)); - + // * Rewrite Section // - Describes how to transform the IR. // - Last statement starts the rewrite. @@ -1009,7 +1009,7 @@ Pattern { // Return a tuple of values. let result = ExtractMultipleResults(op: op<my_dialect.foo>); - // Index the tuple elements by index, or by name. + // Index the tuple elements by index, or by name. replace op<my_dialect.foo> with (result.0, result.1, result.result1); } ``` @@ -1150,7 +1150,7 @@ same name. See the ["type translation"](#native-constraint-type-translations) be detailed information on how PDLL types are converted to native types. In addition to the PDLL arguments, the code block may also access the current `PatternRewriter` using `rewriter`. The result type of the native constraint function is implicitly defined -as a `::mlir::LogicalResult`. +as a `::llvm::LogicalResult`. Taking the constraints defined above as an example, these function would roughly be translated into: @@ -1220,7 +1220,7 @@ was imported: * `Attr` constraints - Imported `Attr` constraints utilize the `storageType` field for native type translation. - + * `Type` constraints - Imported `Type` constraints utilize the `cppClassName` field for native type translation. @@ -1310,7 +1310,7 @@ Pattern { // A pattern that replaces the root operation with another operation. // Note that when an operation is used as the replacement, we can infer its // result types from the input operation. In these cases, the result -// types of replacement operation may be elided. +// types of replacement operation may be elided. Pattern { // Note: In this pattern we also inlined the `root` expression. replace op<my_dialect.foo> with op<my_dialect.bar>; @@ -1385,7 +1385,7 @@ Pattern { // Invoke the rewrite, which returns a tuple of values. let result = CreateRewriteOps(); - // Index the tuple elements by index, or by name. + // Index the tuple elements by index, or by name. replace root with (result.0, result.1, result.result1); } } @@ -1466,7 +1466,7 @@ the C++ PDL API. For example, the rewrite above may be registered as: ```c++ static Operation *buildOpImpl(PDLResultList &results, Value value) { // insert special rewrite logic here. - Operation *resultOp = ...; + Operation *resultOp = ...; return resultOp; } diff --git a/mlir/docs/Tutorials/Toy/Ch-2.md b/mlir/docs/Tutorials/Toy/Ch-2.md index 6801369..b807ee3 100644 --- a/mlir/docs/Tutorials/Toy/Ch-2.md +++ b/mlir/docs/Tutorials/Toy/Ch-2.md @@ -243,7 +243,7 @@ operation. This operation will represent a constant value in the Toy language. This operation takes zero operands, a [dense elements](../../Dialects/Builtin.md/#denseintorfpelementsattr) attribute named `value` to represent the constant value, and returns a single result of -[RankedTensorType](../../Dialects/Builtin.md/#rankedtensortype). An operation class +[RankedTensorType](../../Dialects/Builtin.md/#rankedtensortype). An operation class inherits from the [CRTP](https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern) `mlir::Op` class which also takes some optional [*traits*](../../Traits) to customize its behavior. `Traits` are a mechanism with which we can inject @@ -497,10 +497,10 @@ def ConstantOp : Toy_Op<"constant"> { let results = (outs F64Tensor); // Add additional verification logic to the constant operation. Setting this bit - // to `1` will generate a `::mlir::LogicalResult verify()` declaration on the + // to `1` will generate a `::llvm::LogicalResult verify()` declaration on the // operation class that is called after ODS constructs have been verified, for // example the types of arguments and results. We implement additional verification - // in the definition of this `verify` method in the C++ source file. + // in the definition of this `verify` method in the C++ source file. let hasVerifier = 1; } ``` diff --git a/mlir/docs/Tutorials/Toy/Ch-3.md b/mlir/docs/Tutorials/Toy/Ch-3.md index ac38cbe..4cde64d 100644 --- a/mlir/docs/Tutorials/Toy/Ch-3.md +++ b/mlir/docs/Tutorials/Toy/Ch-3.md @@ -86,7 +86,7 @@ struct SimplifyRedundantTranspose : public mlir::OpRewritePattern<TransposeOp> { /// This method is attempting to match a pattern and rewrite it. The rewriter /// argument is the orchestrator of the sequence of rewrites. It is expected /// to interact with it to perform any changes to the IR from here. - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(TransposeOp op, mlir::PatternRewriter &rewriter) const override { // Look through the input of the current transpose. @@ -128,7 +128,7 @@ similar way to LLVM: pm.addNestedPass<mlir::toy::FuncOp>(mlir::createCanonicalizerPass()); ``` -Finally, we can run `toyc-ch3 test/Examples/Toy/Ch3/transpose_transpose.toy +Finally, we can run `toyc-ch3 test/Examples/Toy/Ch3/transpose_transpose.toy -emit=mlir -opt` and observe our pattern in action: ```mlir @@ -239,7 +239,7 @@ module { } ``` -We can try to run `toyc-ch3 test/Examples/Toy/Ch3/trivial_reshape.toy -emit=mlir +We can try to run `toyc-ch3 test/Examples/Toy/Ch3/trivial_reshape.toy -emit=mlir -opt` and observe our pattern in action: ```mlir diff --git a/mlir/docs/Tutorials/Toy/Ch-5.md b/mlir/docs/Tutorials/Toy/Ch-5.md index 6fe0080..7d4f95e 100644 --- a/mlir/docs/Tutorials/Toy/Ch-5.md +++ b/mlir/docs/Tutorials/Toy/Ch-5.md @@ -113,7 +113,7 @@ struct TransposeOpLowering : public mlir::ConversionPattern { /// Match and rewrite the given `toy.transpose` operation, with the given /// operands that have been remapped from `tensor<...>` to `memref<...>`. - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(mlir::Operation *op, ArrayRef<mlir::Value> operands, mlir::ConversionPatternRewriter &rewriter) const final { auto loc = op->getLoc(); diff --git a/mlir/examples/standalone/lib/Standalone/StandalonePasses.cpp b/mlir/examples/standalone/lib/Standalone/StandalonePasses.cpp index a23d042..8166aa2 100644 --- a/mlir/examples/standalone/lib/Standalone/StandalonePasses.cpp +++ b/mlir/examples/standalone/lib/Standalone/StandalonePasses.cpp @@ -8,7 +8,6 @@ #include "mlir/Dialect/Func/IR/FuncOps.h" #include "mlir/IR/PatternMatch.h" #include "mlir/Rewrite/FrozenRewritePatternSet.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include "Standalone/StandalonePasses.h" diff --git a/mlir/examples/standalone/standalone-translate/standalone-translate.cpp b/mlir/examples/standalone/standalone-translate/standalone-translate.cpp index a4da328..46b9c77 100644 --- a/mlir/examples/standalone/standalone-translate/standalone-translate.cpp +++ b/mlir/examples/standalone/standalone-translate/standalone-translate.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/DialectRegistry.h" #include "mlir/IR/Operation.h" #include "mlir/InitAllTranslations.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Tools/mlir-translate/MlirTranslateMain.h" #include "mlir/Tools/mlir-translate/Translation.h" #include "llvm/Support/raw_ostream.h" @@ -27,7 +26,7 @@ int main(int argc, char **argv) { mlir::TranslateFromMLIRRegistration withdescription( "option", "different from option", [](mlir::Operation *op, llvm::raw_ostream &output) { - return mlir::LogicalResult::success(); + return llvm::LogicalResult::success(); }, [](mlir::DialectRegistry &a) {}); diff --git a/mlir/examples/toy/Ch2/mlir/Dialect.cpp b/mlir/examples/toy/Ch2/mlir/Dialect.cpp index d35bd9f..489f348 100644 --- a/mlir/examples/toy/Ch2/mlir/Dialect.cpp +++ b/mlir/examples/toy/Ch2/mlir/Dialect.cpp @@ -22,7 +22,6 @@ #include "mlir/IR/Value.h" #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/StringRef.h" @@ -142,7 +141,7 @@ void ConstantOp::print(mlir::OpAsmPrinter &printer) { /// Verifier for the constant operation. This corresponds to the /// `let hasVerifier = 1` in the op definition. -mlir::LogicalResult ConstantOp::verify() { +llvm::LogicalResult ConstantOp::verify() { // If the return type of the constant is not an unranked tensor, the shape // must match the shape of the attribute holding the data. auto resultType = llvm::dyn_cast<mlir::RankedTensorType>(getResult().getType()); @@ -257,7 +256,7 @@ void MulOp::print(mlir::OpAsmPrinter &p) { printBinaryOp(p, *this); } // ReturnOp //===----------------------------------------------------------------------===// -mlir::LogicalResult ReturnOp::verify() { +llvm::LogicalResult ReturnOp::verify() { // We know that the parent operation is a function, because of the 'HasParent' // trait attached to the operation definition. auto function = cast<FuncOp>((*this)->getParentOp()); @@ -300,7 +299,7 @@ void TransposeOp::build(mlir::OpBuilder &builder, mlir::OperationState &state, state.addOperands(value); } -mlir::LogicalResult TransposeOp::verify() { +llvm::LogicalResult TransposeOp::verify() { auto inputType = llvm::dyn_cast<RankedTensorType>(getOperand().getType()); auto resultType = llvm::dyn_cast<RankedTensorType>(getType()); if (!inputType || !resultType) diff --git a/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp index 2f0a88f..bf4c099 100644 --- a/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp +++ b/mlir/examples/toy/Ch2/mlir/MLIRGen.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/Block.h" #include "mlir/IR/Diagnostics.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" @@ -105,7 +104,7 @@ private: /// Declare a variable in the current scope, return success if the variable /// wasn't declared yet. - mlir::LogicalResult declare(llvm::StringRef var, mlir::Value value) { + llvm::LogicalResult declare(llvm::StringRef var, mlir::Value value) { if (symbolTable.count(var)) return mlir::failure(); symbolTable.insert(var, value); @@ -225,7 +224,7 @@ private: } /// Emit a return operation. This will return failure if any generation fails. - mlir::LogicalResult mlirGen(ReturnExprAST &ret) { + llvm::LogicalResult mlirGen(ReturnExprAST &ret) { auto location = loc(ret.loc()); // 'return' takes an optional expression, handle that case here. @@ -337,7 +336,7 @@ private: /// Emit a print expression. It emits specific operations for two builtins: /// transpose(x) and print(x). - mlir::LogicalResult mlirGen(PrintExprAST &call) { + llvm::LogicalResult mlirGen(PrintExprAST &call) { auto arg = mlirGen(*call.getArg()); if (!arg) return mlir::failure(); @@ -403,7 +402,7 @@ private: } /// Codegen a list of expression, return failure if one of them hit an error. - mlir::LogicalResult mlirGen(ExprASTList &blockAST) { + llvm::LogicalResult mlirGen(ExprASTList &blockAST) { ScopedHashTableScope<StringRef, mlir::Value> varScope(symbolTable); for (auto &expr : blockAST) { // Specific handling for variable declarations, return statement, and diff --git a/mlir/examples/toy/Ch3/mlir/Dialect.cpp b/mlir/examples/toy/Ch3/mlir/Dialect.cpp index 79d82e5..708855f 100644 --- a/mlir/examples/toy/Ch3/mlir/Dialect.cpp +++ b/mlir/examples/toy/Ch3/mlir/Dialect.cpp @@ -22,7 +22,6 @@ #include "mlir/IR/Value.h" #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/StringRef.h" @@ -142,7 +141,7 @@ void ConstantOp::print(mlir::OpAsmPrinter &printer) { /// Verifier for the constant operation. This corresponds to the /// `let hasVerifier = 1` in the op definition. -mlir::LogicalResult ConstantOp::verify() { +llvm::LogicalResult ConstantOp::verify() { // If the return type of the constant is not an unranked tensor, the shape // must match the shape of the attribute holding the data. auto resultType = llvm::dyn_cast<mlir::RankedTensorType>(getResult().getType()); @@ -257,7 +256,7 @@ void MulOp::print(mlir::OpAsmPrinter &p) { printBinaryOp(p, *this); } // ReturnOp //===----------------------------------------------------------------------===// -mlir::LogicalResult ReturnOp::verify() { +llvm::LogicalResult ReturnOp::verify() { // We know that the parent operation is a function, because of the 'HasParent' // trait attached to the operation definition. auto function = cast<FuncOp>((*this)->getParentOp()); @@ -300,7 +299,7 @@ void TransposeOp::build(mlir::OpBuilder &builder, mlir::OperationState &state, state.addOperands(value); } -mlir::LogicalResult TransposeOp::verify() { +llvm::LogicalResult TransposeOp::verify() { auto inputType = llvm::dyn_cast<RankedTensorType>(getOperand().getType()); auto resultType = llvm::dyn_cast<RankedTensorType>(getType()); if (!inputType || !resultType) diff --git a/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp index 2f0a88f..bf4c099 100644 --- a/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp +++ b/mlir/examples/toy/Ch3/mlir/MLIRGen.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/Block.h" #include "mlir/IR/Diagnostics.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" @@ -105,7 +104,7 @@ private: /// Declare a variable in the current scope, return success if the variable /// wasn't declared yet. - mlir::LogicalResult declare(llvm::StringRef var, mlir::Value value) { + llvm::LogicalResult declare(llvm::StringRef var, mlir::Value value) { if (symbolTable.count(var)) return mlir::failure(); symbolTable.insert(var, value); @@ -225,7 +224,7 @@ private: } /// Emit a return operation. This will return failure if any generation fails. - mlir::LogicalResult mlirGen(ReturnExprAST &ret) { + llvm::LogicalResult mlirGen(ReturnExprAST &ret) { auto location = loc(ret.loc()); // 'return' takes an optional expression, handle that case here. @@ -337,7 +336,7 @@ private: /// Emit a print expression. It emits specific operations for two builtins: /// transpose(x) and print(x). - mlir::LogicalResult mlirGen(PrintExprAST &call) { + llvm::LogicalResult mlirGen(PrintExprAST &call) { auto arg = mlirGen(*call.getArg()); if (!arg) return mlir::failure(); @@ -403,7 +402,7 @@ private: } /// Codegen a list of expression, return failure if one of them hit an error. - mlir::LogicalResult mlirGen(ExprASTList &blockAST) { + llvm::LogicalResult mlirGen(ExprASTList &blockAST) { ScopedHashTableScope<StringRef, mlir::Value> varScope(symbolTable); for (auto &expr : blockAST) { // Specific handling for variable declarations, return statement, and diff --git a/mlir/examples/toy/Ch3/mlir/ToyCombine.cpp b/mlir/examples/toy/Ch3/mlir/ToyCombine.cpp index 3ce35c8..f8397c2 100644 --- a/mlir/examples/toy/Ch3/mlir/ToyCombine.cpp +++ b/mlir/examples/toy/Ch3/mlir/ToyCombine.cpp @@ -14,7 +14,6 @@ #include "mlir/IR/MLIRContext.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/Dialect.h" using namespace mlir; using namespace toy; @@ -36,7 +35,7 @@ struct SimplifyRedundantTranspose : public mlir::OpRewritePattern<TransposeOp> { /// This method attempts to match a pattern and rewrite it. The rewriter /// argument is the orchestrator of the sequence of rewrites. The pattern is /// expected to interact with it to perform any changes to the IR from here. - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(TransposeOp op, mlir::PatternRewriter &rewriter) const override { // Look through the input of the current transpose. diff --git a/mlir/examples/toy/Ch3/toyc.cpp b/mlir/examples/toy/Ch3/toyc.cpp index c2c5f1f..f8aa846 100644 --- a/mlir/examples/toy/Ch3/toyc.cpp +++ b/mlir/examples/toy/Ch3/toyc.cpp @@ -11,7 +11,6 @@ //===----------------------------------------------------------------------===// #include "mlir/IR/Diagnostics.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" #include "toy/Lexer.h" diff --git a/mlir/examples/toy/Ch4/mlir/Dialect.cpp b/mlir/examples/toy/Ch4/mlir/Dialect.cpp index 86a0e1d..6c6cdd9 100644 --- a/mlir/examples/toy/Ch4/mlir/Dialect.cpp +++ b/mlir/examples/toy/Ch4/mlir/Dialect.cpp @@ -23,7 +23,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/InliningUtils.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" @@ -204,7 +203,7 @@ void ConstantOp::print(mlir::OpAsmPrinter &printer) { /// Verifier for the constant operation. This corresponds to the /// `let hasVerifier = 1` in the op definition. -mlir::LogicalResult ConstantOp::verify() { +llvm::LogicalResult ConstantOp::verify() { // If the return type of the constant is not an unranked tensor, the shape // must match the shape of the attribute holding the data. auto resultType = llvm::dyn_cast<mlir::RankedTensorType>(getResult().getType()); @@ -372,7 +371,7 @@ void MulOp::inferShapes() { getResult().setType(getLhs().getType()); } // ReturnOp //===----------------------------------------------------------------------===// -mlir::LogicalResult ReturnOp::verify() { +llvm::LogicalResult ReturnOp::verify() { // We know that the parent operation is a function, because of the 'HasParent' // trait attached to the operation definition. auto function = cast<FuncOp>((*this)->getParentOp()); @@ -421,7 +420,7 @@ void TransposeOp::inferShapes() { getResult().setType(RankedTensorType::get(dims, arrayTy.getElementType())); } -mlir::LogicalResult TransposeOp::verify() { +llvm::LogicalResult TransposeOp::verify() { auto inputType = llvm::dyn_cast<RankedTensorType>(getOperand().getType()); auto resultType = llvm::dyn_cast<RankedTensorType>(getType()); if (!inputType || !resultType) diff --git a/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp index 6c5474a..b56e2f7c 100644 --- a/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp +++ b/mlir/examples/toy/Ch4/mlir/MLIRGen.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/Block.h" #include "mlir/IR/Diagnostics.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" @@ -105,7 +104,7 @@ private: /// Declare a variable in the current scope, return success if the variable /// wasn't declared yet. - mlir::LogicalResult declare(llvm::StringRef var, mlir::Value value) { + llvm::LogicalResult declare(llvm::StringRef var, mlir::Value value) { if (symbolTable.count(var)) return mlir::failure(); symbolTable.insert(var, value); @@ -229,7 +228,7 @@ private: } /// Emit a return operation. This will return failure if any generation fails. - mlir::LogicalResult mlirGen(ReturnExprAST &ret) { + llvm::LogicalResult mlirGen(ReturnExprAST &ret) { auto location = loc(ret.loc()); // 'return' takes an optional expression, handle that case here. @@ -341,7 +340,7 @@ private: /// Emit a print expression. It emits specific operations for two builtins: /// transpose(x) and print(x). - mlir::LogicalResult mlirGen(PrintExprAST &call) { + llvm::LogicalResult mlirGen(PrintExprAST &call) { auto arg = mlirGen(*call.getArg()); if (!arg) return mlir::failure(); @@ -407,7 +406,7 @@ private: } /// Codegen a list of expression, return failure if one of them hit an error. - mlir::LogicalResult mlirGen(ExprASTList &blockAST) { + llvm::LogicalResult mlirGen(ExprASTList &blockAST) { ScopedHashTableScope<StringRef, mlir::Value> varScope(symbolTable); for (auto &expr : blockAST) { // Specific handling for variable declarations, return statement, and diff --git a/mlir/examples/toy/Ch4/mlir/ToyCombine.cpp b/mlir/examples/toy/Ch4/mlir/ToyCombine.cpp index 3ce35c8..f8397c2 100644 --- a/mlir/examples/toy/Ch4/mlir/ToyCombine.cpp +++ b/mlir/examples/toy/Ch4/mlir/ToyCombine.cpp @@ -14,7 +14,6 @@ #include "mlir/IR/MLIRContext.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/Dialect.h" using namespace mlir; using namespace toy; @@ -36,7 +35,7 @@ struct SimplifyRedundantTranspose : public mlir::OpRewritePattern<TransposeOp> { /// This method attempts to match a pattern and rewrite it. The rewriter /// argument is the orchestrator of the sequence of rewrites. The pattern is /// expected to interact with it to perform any changes to the IR from here. - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(TransposeOp op, mlir::PatternRewriter &rewriter) const override { // Look through the input of the current transpose. diff --git a/mlir/examples/toy/Ch4/toyc.cpp b/mlir/examples/toy/Ch4/toyc.cpp index a1534ed..ae02bc4 100644 --- a/mlir/examples/toy/Ch4/toyc.cpp +++ b/mlir/examples/toy/Ch4/toyc.cpp @@ -11,7 +11,6 @@ //===----------------------------------------------------------------------===// #include "mlir/IR/Diagnostics.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" #include "toy/Lexer.h" diff --git a/mlir/examples/toy/Ch5/mlir/Dialect.cpp b/mlir/examples/toy/Ch5/mlir/Dialect.cpp index c587dd2..72072f9 100644 --- a/mlir/examples/toy/Ch5/mlir/Dialect.cpp +++ b/mlir/examples/toy/Ch5/mlir/Dialect.cpp @@ -23,7 +23,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/InliningUtils.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" @@ -204,7 +203,7 @@ void ConstantOp::print(mlir::OpAsmPrinter &printer) { /// Verifier for the constant operation. This corresponds to the /// `let hasVerifier = 1` in the op definition. -mlir::LogicalResult ConstantOp::verify() { +llvm::LogicalResult ConstantOp::verify() { // If the return type of the constant is not an unranked tensor, the shape // must match the shape of the attribute holding the data. auto resultType = llvm::dyn_cast<mlir::RankedTensorType>(getResult().getType()); @@ -372,7 +371,7 @@ void MulOp::inferShapes() { getResult().setType(getLhs().getType()); } // ReturnOp //===----------------------------------------------------------------------===// -mlir::LogicalResult ReturnOp::verify() { +llvm::LogicalResult ReturnOp::verify() { // We know that the parent operation is a function, because of the 'HasParent' // trait attached to the operation definition. auto function = cast<FuncOp>((*this)->getParentOp()); @@ -421,7 +420,7 @@ void TransposeOp::inferShapes() { getResult().setType(RankedTensorType::get(dims, arrayTy.getElementType())); } -mlir::LogicalResult TransposeOp::verify() { +llvm::LogicalResult TransposeOp::verify() { auto inputType = llvm::dyn_cast<RankedTensorType>(getOperand().getType()); auto resultType = llvm::dyn_cast<RankedTensorType>(getType()); if (!inputType || !resultType) diff --git a/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp b/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp index bded615..7413214 100644 --- a/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp +++ b/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp @@ -21,7 +21,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/ValueRange.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "toy/Dialect.h" #include "toy/Passes.h" diff --git a/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp index 6c5474a..b56e2f7c 100644 --- a/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp +++ b/mlir/examples/toy/Ch5/mlir/MLIRGen.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/Block.h" #include "mlir/IR/Diagnostics.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" @@ -105,7 +104,7 @@ private: /// Declare a variable in the current scope, return success if the variable /// wasn't declared yet. - mlir::LogicalResult declare(llvm::StringRef var, mlir::Value value) { + llvm::LogicalResult declare(llvm::StringRef var, mlir::Value value) { if (symbolTable.count(var)) return mlir::failure(); symbolTable.insert(var, value); @@ -229,7 +228,7 @@ private: } /// Emit a return operation. This will return failure if any generation fails. - mlir::LogicalResult mlirGen(ReturnExprAST &ret) { + llvm::LogicalResult mlirGen(ReturnExprAST &ret) { auto location = loc(ret.loc()); // 'return' takes an optional expression, handle that case here. @@ -341,7 +340,7 @@ private: /// Emit a print expression. It emits specific operations for two builtins: /// transpose(x) and print(x). - mlir::LogicalResult mlirGen(PrintExprAST &call) { + llvm::LogicalResult mlirGen(PrintExprAST &call) { auto arg = mlirGen(*call.getArg()); if (!arg) return mlir::failure(); @@ -407,7 +406,7 @@ private: } /// Codegen a list of expression, return failure if one of them hit an error. - mlir::LogicalResult mlirGen(ExprASTList &blockAST) { + llvm::LogicalResult mlirGen(ExprASTList &blockAST) { ScopedHashTableScope<StringRef, mlir::Value> varScope(symbolTable); for (auto &expr : blockAST) { // Specific handling for variable declarations, return statement, and diff --git a/mlir/examples/toy/Ch5/mlir/ToyCombine.cpp b/mlir/examples/toy/Ch5/mlir/ToyCombine.cpp index 3ce35c8..f8397c2 100644 --- a/mlir/examples/toy/Ch5/mlir/ToyCombine.cpp +++ b/mlir/examples/toy/Ch5/mlir/ToyCombine.cpp @@ -14,7 +14,6 @@ #include "mlir/IR/MLIRContext.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/Dialect.h" using namespace mlir; using namespace toy; @@ -36,7 +35,7 @@ struct SimplifyRedundantTranspose : public mlir::OpRewritePattern<TransposeOp> { /// This method attempts to match a pattern and rewrite it. The rewriter /// argument is the orchestrator of the sequence of rewrites. The pattern is /// expected to interact with it to perform any changes to the IR from here. - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(TransposeOp op, mlir::PatternRewriter &rewriter) const override { // Look through the input of the current transpose. diff --git a/mlir/examples/toy/Ch5/toyc.cpp b/mlir/examples/toy/Ch5/toyc.cpp index 4eb6fde..6a0c631 100644 --- a/mlir/examples/toy/Ch5/toyc.cpp +++ b/mlir/examples/toy/Ch5/toyc.cpp @@ -12,7 +12,6 @@ #include "mlir/Dialect/Func/Extensions/AllExtensions.h" #include "mlir/IR/Diagnostics.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" #include "toy/Lexer.h" diff --git a/mlir/examples/toy/Ch6/mlir/Dialect.cpp b/mlir/examples/toy/Ch6/mlir/Dialect.cpp index c587dd2..72072f9 100644 --- a/mlir/examples/toy/Ch6/mlir/Dialect.cpp +++ b/mlir/examples/toy/Ch6/mlir/Dialect.cpp @@ -23,7 +23,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/InliningUtils.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" @@ -204,7 +203,7 @@ void ConstantOp::print(mlir::OpAsmPrinter &printer) { /// Verifier for the constant operation. This corresponds to the /// `let hasVerifier = 1` in the op definition. -mlir::LogicalResult ConstantOp::verify() { +llvm::LogicalResult ConstantOp::verify() { // If the return type of the constant is not an unranked tensor, the shape // must match the shape of the attribute holding the data. auto resultType = llvm::dyn_cast<mlir::RankedTensorType>(getResult().getType()); @@ -372,7 +371,7 @@ void MulOp::inferShapes() { getResult().setType(getLhs().getType()); } // ReturnOp //===----------------------------------------------------------------------===// -mlir::LogicalResult ReturnOp::verify() { +llvm::LogicalResult ReturnOp::verify() { // We know that the parent operation is a function, because of the 'HasParent' // trait attached to the operation definition. auto function = cast<FuncOp>((*this)->getParentOp()); @@ -421,7 +420,7 @@ void TransposeOp::inferShapes() { getResult().setType(RankedTensorType::get(dims, arrayTy.getElementType())); } -mlir::LogicalResult TransposeOp::verify() { +llvm::LogicalResult TransposeOp::verify() { auto inputType = llvm::dyn_cast<RankedTensorType>(getOperand().getType()); auto resultType = llvm::dyn_cast<RankedTensorType>(getType()); if (!inputType || !resultType) diff --git a/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp b/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp index bded615..7413214 100644 --- a/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp +++ b/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp @@ -21,7 +21,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/ValueRange.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "toy/Dialect.h" #include "toy/Passes.h" diff --git a/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp b/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp index f91d880..3ad70e7 100644 --- a/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp +++ b/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp @@ -28,7 +28,6 @@ #include "mlir/IR/BuiltinOps.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "toy/Dialect.h" #include "toy/Passes.h" diff --git a/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp index 6c5474a..b56e2f7c 100644 --- a/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp +++ b/mlir/examples/toy/Ch6/mlir/MLIRGen.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/Block.h" #include "mlir/IR/Diagnostics.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" @@ -105,7 +104,7 @@ private: /// Declare a variable in the current scope, return success if the variable /// wasn't declared yet. - mlir::LogicalResult declare(llvm::StringRef var, mlir::Value value) { + llvm::LogicalResult declare(llvm::StringRef var, mlir::Value value) { if (symbolTable.count(var)) return mlir::failure(); symbolTable.insert(var, value); @@ -229,7 +228,7 @@ private: } /// Emit a return operation. This will return failure if any generation fails. - mlir::LogicalResult mlirGen(ReturnExprAST &ret) { + llvm::LogicalResult mlirGen(ReturnExprAST &ret) { auto location = loc(ret.loc()); // 'return' takes an optional expression, handle that case here. @@ -341,7 +340,7 @@ private: /// Emit a print expression. It emits specific operations for two builtins: /// transpose(x) and print(x). - mlir::LogicalResult mlirGen(PrintExprAST &call) { + llvm::LogicalResult mlirGen(PrintExprAST &call) { auto arg = mlirGen(*call.getArg()); if (!arg) return mlir::failure(); @@ -407,7 +406,7 @@ private: } /// Codegen a list of expression, return failure if one of them hit an error. - mlir::LogicalResult mlirGen(ExprASTList &blockAST) { + llvm::LogicalResult mlirGen(ExprASTList &blockAST) { ScopedHashTableScope<StringRef, mlir::Value> varScope(symbolTable); for (auto &expr : blockAST) { // Specific handling for variable declarations, return statement, and diff --git a/mlir/examples/toy/Ch6/mlir/ToyCombine.cpp b/mlir/examples/toy/Ch6/mlir/ToyCombine.cpp index 3ce35c8..f8397c2 100644 --- a/mlir/examples/toy/Ch6/mlir/ToyCombine.cpp +++ b/mlir/examples/toy/Ch6/mlir/ToyCombine.cpp @@ -14,7 +14,6 @@ #include "mlir/IR/MLIRContext.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/Dialect.h" using namespace mlir; using namespace toy; @@ -36,7 +35,7 @@ struct SimplifyRedundantTranspose : public mlir::OpRewritePattern<TransposeOp> { /// This method attempts to match a pattern and rewrite it. The rewriter /// argument is the orchestrator of the sequence of rewrites. The pattern is /// expected to interact with it to perform any changes to the IR from here. - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(TransposeOp op, mlir::PatternRewriter &rewriter) const override { // Look through the input of the current transpose. diff --git a/mlir/examples/toy/Ch6/toyc.cpp b/mlir/examples/toy/Ch6/toyc.cpp index ddc0c25..c244b31 100644 --- a/mlir/examples/toy/Ch6/toyc.cpp +++ b/mlir/examples/toy/Ch6/toyc.cpp @@ -12,7 +12,6 @@ #include "mlir/Dialect/Func/Extensions/AllExtensions.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" #include "toy/Lexer.h" diff --git a/mlir/examples/toy/Ch7/mlir/Dialect.cpp b/mlir/examples/toy/Ch7/mlir/Dialect.cpp index b268b1e..7e030ff 100644 --- a/mlir/examples/toy/Ch7/mlir/Dialect.cpp +++ b/mlir/examples/toy/Ch7/mlir/Dialect.cpp @@ -27,7 +27,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/InliningUtils.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/Hashing.h" @@ -195,7 +194,7 @@ void ConstantOp::print(mlir::OpAsmPrinter &printer) { } /// Verify that the given attribute value is valid for the given type. -static mlir::LogicalResult verifyConstantForType(mlir::Type type, +static llvm::LogicalResult verifyConstantForType(mlir::Type type, mlir::Attribute opaqueValue, mlir::Operation *op) { if (llvm::isa<mlir::TensorType>(type)) { @@ -252,11 +251,11 @@ static mlir::LogicalResult verifyConstantForType(mlir::Type type, /// Verifier for the constant operation. This corresponds to the `::verify(...)` /// in the op definition. -mlir::LogicalResult ConstantOp::verify() { +llvm::LogicalResult ConstantOp::verify() { return verifyConstantForType(getResult().getType(), getValue(), *this); } -mlir::LogicalResult StructConstantOp::verify() { +llvm::LogicalResult StructConstantOp::verify() { return verifyConstantForType(getResult().getType(), getValue(), *this); } @@ -406,7 +405,7 @@ void MulOp::inferShapes() { getResult().setType(getLhs().getType()); } // ReturnOp //===----------------------------------------------------------------------===// -mlir::LogicalResult ReturnOp::verify() { +llvm::LogicalResult ReturnOp::verify() { // We know that the parent operation is a function, because of the 'HasParent' // trait attached to the operation definition. auto function = cast<FuncOp>((*this)->getParentOp()); @@ -454,7 +453,7 @@ void StructAccessOp::build(mlir::OpBuilder &b, mlir::OperationState &state, build(b, state, resultType, input, b.getI64IntegerAttr(index)); } -mlir::LogicalResult StructAccessOp::verify() { +llvm::LogicalResult StructAccessOp::verify() { StructType structTy = llvm::cast<StructType>(getInput().getType()); size_t indexValue = getIndex(); if (indexValue >= structTy.getNumElementTypes()) @@ -483,7 +482,7 @@ void TransposeOp::inferShapes() { getResult().setType(RankedTensorType::get(dims, arrayTy.getElementType())); } -mlir::LogicalResult TransposeOp::verify() { +llvm::LogicalResult TransposeOp::verify() { auto inputType = llvm::dyn_cast<RankedTensorType>(getOperand().getType()); auto resultType = llvm::dyn_cast<RankedTensorType>(getType()); if (!inputType || !resultType) diff --git a/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp b/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp index bded615..7413214 100644 --- a/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp +++ b/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp @@ -21,7 +21,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/ValueRange.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "toy/Dialect.h" #include "toy/Passes.h" diff --git a/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp b/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp index f91d880..3ad70e7 100644 --- a/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp +++ b/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp @@ -28,7 +28,6 @@ #include "mlir/IR/BuiltinOps.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "toy/Dialect.h" #include "toy/Passes.h" diff --git a/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp b/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp index 0f8e8df..090e5ff 100644 --- a/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp +++ b/mlir/examples/toy/Ch7/mlir/MLIRGen.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/Block.h" #include "mlir/IR/Diagnostics.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" @@ -133,7 +132,7 @@ private: /// Declare a variable in the current scope, return success if the variable /// wasn't declared yet. - mlir::LogicalResult declare(VarDeclExprAST &var, mlir::Value value) { + llvm::LogicalResult declare(VarDeclExprAST &var, mlir::Value value) { if (symbolTable.count(var.getName())) return mlir::failure(); symbolTable.insert(var.getName(), {value, &var}); @@ -141,7 +140,7 @@ private: } /// Create an MLIR type for the given struct. - mlir::LogicalResult mlirGen(StructAST &str) { + llvm::LogicalResult mlirGen(StructAST &str) { if (structMap.count(str.getName())) return emitError(loc(str.loc())) << "error: struct type with name `" << str.getName() << "' already exists"; @@ -368,7 +367,7 @@ private: } /// Emit a return operation. This will return failure if any generation fails. - mlir::LogicalResult mlirGen(ReturnExprAST &ret) { + llvm::LogicalResult mlirGen(ReturnExprAST &ret) { auto location = loc(ret.loc()); // 'return' takes an optional expression, handle that case here. @@ -542,7 +541,7 @@ private: /// Emit a print expression. It emits specific operations for two builtins: /// transpose(x) and print(x). - mlir::LogicalResult mlirGen(PrintExprAST &call) { + llvm::LogicalResult mlirGen(PrintExprAST &call) { auto arg = mlirGen(*call.getArg()); if (!arg) return mlir::failure(); @@ -626,7 +625,7 @@ private: } /// Codegen a list of expression, return failure if one of them hit an error. - mlir::LogicalResult mlirGen(ExprASTList &blockAST) { + llvm::LogicalResult mlirGen(ExprASTList &blockAST) { SymbolTableScopeT varScope(symbolTable); for (auto &expr : blockAST) { // Specific handling for variable declarations, return statement, and diff --git a/mlir/examples/toy/Ch7/mlir/ToyCombine.cpp b/mlir/examples/toy/Ch7/mlir/ToyCombine.cpp index 72f5e4b..1d8cf74 100644 --- a/mlir/examples/toy/Ch7/mlir/ToyCombine.cpp +++ b/mlir/examples/toy/Ch7/mlir/ToyCombine.cpp @@ -16,7 +16,6 @@ #include "mlir/IR/OpDefinition.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "toy/Dialect.h" #include "llvm/Support/Casting.h" #include <cstddef> @@ -57,7 +56,7 @@ struct SimplifyRedundantTranspose : public mlir::OpRewritePattern<TransposeOp> { /// This method attempts to match a pattern and rewrite it. The rewriter /// argument is the orchestrator of the sequence of rewrites. The pattern is /// expected to interact with it to perform any changes to the IR from here. - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(TransposeOp op, mlir::PatternRewriter &rewriter) const override { // Look through the input of the current transpose. diff --git a/mlir/examples/toy/Ch7/toyc.cpp b/mlir/examples/toy/Ch7/toyc.cpp index 5eb40b7..fea5679 100644 --- a/mlir/examples/toy/Ch7/toyc.cpp +++ b/mlir/examples/toy/Ch7/toyc.cpp @@ -12,7 +12,6 @@ #include "mlir/Dialect/Func/Extensions/AllExtensions.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" -#include "mlir/Support/LogicalResult.h" #include "toy/AST.h" #include "toy/Dialect.h" #include "toy/Lexer.h" diff --git a/mlir/examples/transform-opt/mlir-transform-opt.cpp b/mlir/examples/transform-opt/mlir-transform-opt.cpp index 41a17f1..65615fb 100644 --- a/mlir/examples/transform-opt/mlir-transform-opt.cpp +++ b/mlir/examples/transform-opt/mlir-transform-opt.cpp @@ -118,7 +118,7 @@ public: DiagnosticHandlerWrapper &operator=(DiagnosticHandlerWrapper &&) = default; /// Verifies the captured "expected-*" diagnostics if required. - mlir::LogicalResult verify() const { + llvm::LogicalResult verify() const { if (auto *ptr = handler.dyn_cast<mlir::SourceMgrDiagnosticVerifierHandler *>()) { return ptr->verify(); @@ -192,7 +192,7 @@ public: /// If diagnostic message verification has been requested upon construction of /// this source manager, performs the verification, reports errors and returns /// the result of the verification. Otherwise passes through the given value. - mlir::LogicalResult checkResult(mlir::LogicalResult result) { + llvm::LogicalResult checkResult(llvm::LogicalResult result) { resultChecked = true; if (!verifyDiagnostics) return result; @@ -222,7 +222,7 @@ private: /// Trivial wrapper around `applyTransforms` that doesn't support extra mapping /// and doesn't enforce the entry point transform ops being top-level. -static mlir::LogicalResult +static llvm::LogicalResult applyTransforms(mlir::Operation *payloadRoot, mlir::transform::TransformOpInterface transformRoot, const mlir::transform::TransformOptions &options) { @@ -237,7 +237,7 @@ applyTransforms(mlir::Operation *payloadRoot, /// application is successful, prints the transformed input buffer into the /// given output stream. Additional configuration options are derived from /// command-line options. -static mlir::LogicalResult processPayloadBuffer( +static llvm::LogicalResult processPayloadBuffer( raw_ostream &os, std::unique_ptr<MemoryBuffer> inputBuffer, std::unique_ptr<llvm::MemoryBuffer> transformBuffer, MutableArrayRef<std::unique_ptr<MemoryBuffer>> transformLibraries, @@ -309,7 +309,7 @@ static mlir::LogicalResult processPayloadBuffer( } /// Tool entry point. -static mlir::LogicalResult runMain(int argc, char **argv) { +static llvm::LogicalResult runMain(int argc, char **argv) { // Register all upstream dialects and extensions. Specific uses are advised // not to register all dialects indiscriminately but rather hand-pick what is // necessary for their use case. diff --git a/mlir/examples/transform/Ch4/lib/MyExtension.cpp b/mlir/examples/transform/Ch4/lib/MyExtension.cpp index 0790736..38c8ca1 100644 --- a/mlir/examples/transform/Ch4/lib/MyExtension.cpp +++ b/mlir/examples/transform/Ch4/lib/MyExtension.cpp @@ -166,7 +166,7 @@ void mlir::transform::HasOperandSatisfyingOp::getEffects( // Verify well-formedness of the operation and emit diagnostics if it is // ill-formed. -mlir::LogicalResult mlir::transform::HasOperandSatisfyingOp::verify() { +llvm::LogicalResult mlir::transform::HasOperandSatisfyingOp::verify() { mlir::Block &bodyBlock = getBody().front(); if (bodyBlock.getNumArguments() != 1 || !isa<TransformValueHandleTypeInterface>( diff --git a/mlir/include/mlir/Analysis/FlatLinearValueConstraints.h b/mlir/include/mlir/Analysis/FlatLinearValueConstraints.h index cc6ab64..63305e0 100644 --- a/mlir/include/mlir/Analysis/FlatLinearValueConstraints.h +++ b/mlir/include/mlir/Analysis/FlatLinearValueConstraints.h @@ -13,7 +13,6 @@ #include "mlir/Analysis/Presburger/Matrix.h" #include "mlir/IR/AffineExpr.h" #include "mlir/IR/OpDefinition.h" -#include "mlir/Support/LogicalResult.h" #include <optional> namespace mlir { diff --git a/mlir/include/mlir/Bytecode/BytecodeImplementation.h b/mlir/include/mlir/Bytecode/BytecodeImplementation.h index bc91318c..0ddc531 100644 --- a/mlir/include/mlir/Bytecode/BytecodeImplementation.h +++ b/mlir/include/mlir/Bytecode/BytecodeImplementation.h @@ -19,7 +19,6 @@ #include "mlir/IR/Dialect.h" #include "mlir/IR/DialectInterface.h" #include "mlir/IR/OpImplementation.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/Twine.h" diff --git a/mlir/include/mlir/Bytecode/BytecodeOpInterface.h b/mlir/include/mlir/Bytecode/BytecodeOpInterface.h index 10f5be0..c9c6060 100644 --- a/mlir/include/mlir/Bytecode/BytecodeOpInterface.h +++ b/mlir/include/mlir/Bytecode/BytecodeOpInterface.h @@ -18,7 +18,6 @@ #include "mlir/Bytecode/BytecodeReader.h" #include "mlir/Bytecode/BytecodeWriter.h" #include "mlir/IR/OpDefinition.h" -#include "mlir/Support/LogicalResult.h" /// Include the generated interface declarations. #include "mlir/Bytecode/BytecodeOpInterface.h.inc" diff --git a/mlir/include/mlir/Bytecode/BytecodeReader.h b/mlir/include/mlir/Bytecode/BytecodeReader.h index 9f26506..19f74a0 100644 --- a/mlir/include/mlir/Bytecode/BytecodeReader.h +++ b/mlir/include/mlir/Bytecode/BytecodeReader.h @@ -15,7 +15,6 @@ #include "mlir/IR/AsmState.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include <functional> #include <memory> diff --git a/mlir/include/mlir/Bytecode/BytecodeReaderConfig.h b/mlir/include/mlir/Bytecode/BytecodeReaderConfig.h index d623d0d..47be732 100644 --- a/mlir/include/mlir/Bytecode/BytecodeReaderConfig.h +++ b/mlir/include/mlir/Bytecode/BytecodeReaderConfig.h @@ -14,7 +14,6 @@ #define MLIR_BYTECODE_BYTECODEREADERCONFIG_H #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/SmallVector.h" #include "llvm/ADT/StringRef.h" diff --git a/mlir/include/mlir/CAPI/Support.h b/mlir/include/mlir/CAPI/Support.h index 6227452..89a4603 100644 --- a/mlir/include/mlir/CAPI/Support.h +++ b/mlir/include/mlir/CAPI/Support.h @@ -17,9 +17,9 @@ #include "mlir-c/Support.h" #include "mlir/CAPI/Wrap.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "llvm/ADT/StringRef.h" +#include "llvm/Support/LogicalResult.h" namespace llvm { class ThreadPoolInterface; @@ -35,13 +35,13 @@ inline llvm::StringRef unwrap(MlirStringRef ref) { return llvm::StringRef(ref.data, ref.length); } -inline MlirLogicalResult wrap(mlir::LogicalResult res) { +inline MlirLogicalResult wrap(llvm::LogicalResult res) { if (mlir::succeeded(res)) return mlirLogicalResultSuccess(); return mlirLogicalResultFailure(); } -inline mlir::LogicalResult unwrap(MlirLogicalResult res) { +inline llvm::LogicalResult unwrap(MlirLogicalResult res) { return mlir::success(mlirLogicalResultIsSuccess(res)); } diff --git a/mlir/include/mlir/Conversion/AffineToStandard/AffineToStandard.h b/mlir/include/mlir/Conversion/AffineToStandard/AffineToStandard.h index 3850a00..96ee4f0 100644 --- a/mlir/include/mlir/Conversion/AffineToStandard/AffineToStandard.h +++ b/mlir/include/mlir/Conversion/AffineToStandard/AffineToStandard.h @@ -13,7 +13,6 @@ namespace mlir { class Location; -struct LogicalResult; class OpBuilder; class Pass; class RewritePattern; diff --git a/mlir/include/mlir/Conversion/ConvertToLLVM/ToLLVMInterface.h b/mlir/include/mlir/Conversion/ConvertToLLVM/ToLLVMInterface.h index 424ab38..00aeed9 100644 --- a/mlir/include/mlir/Conversion/ConvertToLLVM/ToLLVMInterface.h +++ b/mlir/include/mlir/Conversion/ConvertToLLVM/ToLLVMInterface.h @@ -11,7 +11,6 @@ #include "mlir/IR/DialectInterface.h" #include "mlir/IR/MLIRContext.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { class ConversionTarget; diff --git a/mlir/include/mlir/Conversion/FuncToLLVM/ConvertFuncToLLVM.h b/mlir/include/mlir/Conversion/FuncToLLVM/ConvertFuncToLLVM.h index 107718e..76eb3b5 100644 --- a/mlir/include/mlir/Conversion/FuncToLLVM/ConvertFuncToLLVM.h +++ b/mlir/include/mlir/Conversion/FuncToLLVM/ConvertFuncToLLVM.h @@ -15,7 +15,6 @@ #define MLIR_CONVERSION_FUNCTOLLVM_CONVERTFUNCTOLLVM_H #include "mlir/Interfaces/FunctionInterfaces.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { diff --git a/mlir/include/mlir/Conversion/GPUCommon/GPUCommonPass.h b/mlir/include/mlir/Conversion/GPUCommon/GPUCommonPass.h index 2d5e9d2..5f40315 100644 --- a/mlir/include/mlir/Conversion/GPUCommon/GPUCommonPass.h +++ b/mlir/include/mlir/Conversion/GPUCommon/GPUCommonPass.h @@ -26,7 +26,6 @@ namespace mlir { class LLVMTypeConverter; class Location; -struct LogicalResult; class ModuleOp; class Operation; class RewritePatternSet; diff --git a/mlir/include/mlir/Conversion/SCFToGPU/SCFToGPU.h b/mlir/include/mlir/Conversion/SCFToGPU/SCFToGPU.h index a180a12..69e3f70 100644 --- a/mlir/include/mlir/Conversion/SCFToGPU/SCFToGPU.h +++ b/mlir/include/mlir/Conversion/SCFToGPU/SCFToGPU.h @@ -12,7 +12,6 @@ namespace mlir { class ConversionTarget; -struct LogicalResult; class MLIRContext; class Value; class Operation; diff --git a/mlir/include/mlir/Conversion/VectorToGPU/VectorToGPU.h b/mlir/include/mlir/Conversion/VectorToGPU/VectorToGPU.h index 3530eb5..10467e6 100644 --- a/mlir/include/mlir/Conversion/VectorToGPU/VectorToGPU.h +++ b/mlir/include/mlir/Conversion/VectorToGPU/VectorToGPU.h @@ -12,7 +12,6 @@ #include "mlir/IR/PatternMatch.h" namespace mlir { -struct LogicalResult; class MLIRContext; class Pass; class RewritePatternSet; diff --git a/mlir/include/mlir/Debug/CLOptionsSetup.h b/mlir/include/mlir/Debug/CLOptionsSetup.h index a30ffe0..e7e9727 100644 --- a/mlir/include/mlir/Debug/CLOptionsSetup.h +++ b/mlir/include/mlir/Debug/CLOptionsSetup.h @@ -10,7 +10,6 @@ #define MLIR_DEBUG_CLOPTIONSSETUP_H #include "mlir/Debug/BreakpointManagers/FileLineColLocBreakpointManager.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/StringRef.h" #include <memory> diff --git a/mlir/include/mlir/Dialect/AMDGPU/Utils/Chipset.h b/mlir/include/mlir/Dialect/AMDGPU/Utils/Chipset.h index 5031dc4..38e0ebe 100644 --- a/mlir/include/mlir/Dialect/AMDGPU/Utils/Chipset.h +++ b/mlir/include/mlir/Dialect/AMDGPU/Utils/Chipset.h @@ -8,7 +8,7 @@ #ifndef MLIR_DIALECT_AMDGPU_UTILS_CHIPSET_H_ #define MLIR_DIALECT_AMDGPU_UTILS_CHIPSET_H_ -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" namespace mlir { namespace amdgpu { diff --git a/mlir/include/mlir/Dialect/Affine/Analysis/AffineStructures.h b/mlir/include/mlir/Dialect/Affine/Analysis/AffineStructures.h index c9d3dc1..707eec7 100644 --- a/mlir/include/mlir/Dialect/Affine/Analysis/AffineStructures.h +++ b/mlir/include/mlir/Dialect/Affine/Analysis/AffineStructures.h @@ -18,7 +18,6 @@ #include "mlir/Analysis/Presburger/Matrix.h" #include "mlir/IR/AffineExpr.h" #include "mlir/IR/OpDefinition.h" -#include "mlir/Support/LogicalResult.h" #include <optional> namespace mlir { diff --git a/mlir/include/mlir/Dialect/Affine/IR/ValueBoundsOpInterfaceImpl.h b/mlir/include/mlir/Dialect/Affine/IR/ValueBoundsOpInterfaceImpl.h index 6e617ef4..451c466 100644 --- a/mlir/include/mlir/Dialect/Affine/IR/ValueBoundsOpInterfaceImpl.h +++ b/mlir/include/mlir/Dialect/Affine/IR/ValueBoundsOpInterfaceImpl.h @@ -9,7 +9,7 @@ #ifndef MLIR_DIALECT_AFFINE_IR_VALUEBOUNDSOPINTERFACEIMPL_H #define MLIR_DIALECT_AFFINE_IR_VALUEBOUNDSOPINTERFACEIMPL_H -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" namespace mlir { class DialectRegistry; diff --git a/mlir/include/mlir/Dialect/Affine/LoopUtils.h b/mlir/include/mlir/Dialect/Affine/LoopUtils.h index d143954..99c500c 100644 --- a/mlir/include/mlir/Dialect/Affine/LoopUtils.h +++ b/mlir/include/mlir/Dialect/Affine/LoopUtils.h @@ -17,7 +17,6 @@ #include "mlir/IR/Block.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/RegionUtils.h" #include <optional> diff --git a/mlir/include/mlir/Dialect/Affine/Transforms/Transforms.h b/mlir/include/mlir/Dialect/Affine/Transforms/Transforms.h index 1ea7375..b244d37 100644 --- a/mlir/include/mlir/Dialect/Affine/Transforms/Transforms.h +++ b/mlir/include/mlir/Dialect/Affine/Transforms/Transforms.h @@ -16,7 +16,6 @@ #include "mlir/Interfaces/ValueBoundsOpInterface.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { class AffineMap; diff --git a/mlir/include/mlir/Dialect/Affine/Utils.h b/mlir/include/mlir/Dialect/Affine/Utils.h index 7f25db0..9a2767e 100644 --- a/mlir/include/mlir/Dialect/Affine/Utils.h +++ b/mlir/include/mlir/Dialect/Affine/Utils.h @@ -33,8 +33,6 @@ namespace memref { class AllocOp; } // namespace memref -struct LogicalResult; - namespace affine { class AffineForOp; class AffineIfOp; diff --git a/mlir/include/mlir/Dialect/Arith/Transforms/Transforms.h b/mlir/include/mlir/Dialect/Arith/Transforms/Transforms.h index bbc7e5d..8d6c68c 100644 --- a/mlir/include/mlir/Dialect/Arith/Transforms/Transforms.h +++ b/mlir/include/mlir/Dialect/Arith/Transforms/Transforms.h @@ -10,7 +10,6 @@ #define MLIR_DIALECT_ARITH_TRANSFORMS_TRANSFORMS_H #include "mlir/Interfaces/ValueBoundsOpInterface.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { class Location; diff --git a/mlir/include/mlir/Dialect/Bufferization/IR/BufferizableOpInterface.td b/mlir/include/mlir/Dialect/Bufferization/IR/BufferizableOpInterface.td index 007c05a..80cd13d 100644 --- a/mlir/include/mlir/Dialect/Bufferization/IR/BufferizableOpInterface.td +++ b/mlir/include/mlir/Dialect/Bufferization/IR/BufferizableOpInterface.td @@ -378,7 +378,7 @@ def BufferizableOpInterface : OpInterface<"BufferizableOpInterface"> { This method can query analysis information from the given analysis state. }], - /*retType=*/"::mlir::LogicalResult", + /*retType=*/"::llvm::LogicalResult", /*methodName=*/"resolveConflicts", /*args=*/(ins "::mlir::RewriterBase &":$rewriter, "const ::mlir::bufferization::AnalysisState &":$state), @@ -423,7 +423,7 @@ def BufferizableOpInterface : OpInterface<"BufferizableOpInterface"> { suggestion to make sure IR is valid at every point in time and could be done differently). }], - /*retType=*/"::mlir::LogicalResult", + /*retType=*/"::llvm::LogicalResult", /*methodName=*/"bufferize", /*args=*/(ins "::mlir::RewriterBase &":$rewriter, "const ::mlir::bufferization::BufferizationOptions &":$options), @@ -490,7 +490,7 @@ def BufferizableOpInterface : OpInterface<"BufferizableOpInterface"> { This method can be used to check expected invariants and limitations of the current bufferization implementation. }], - /*retType=*/"::mlir::LogicalResult", + /*retType=*/"::llvm::LogicalResult", /*methodName=*/"verifyAnalysis", /*args=*/(ins "const ::mlir::bufferization::AnalysisState &":$state), /*methodBody=*/"", @@ -613,7 +613,7 @@ def BufferizableOpInterface : OpInterface<"BufferizableOpInterface"> { let extraClassDeclaration = [{ /// Resolve out-of-place tensor OpOperands with explicit allocations in the /// form of `bufferization.alloc_tensor` ops. - ::mlir::LogicalResult resolveTensorOpOperandConflicts( + ::llvm::LogicalResult resolveTensorOpOperandConflicts( ::mlir::RewriterBase &rewriter, const ::mlir::bufferization::AnalysisState &state); diff --git a/mlir/include/mlir/Dialect/Bufferization/Transforms/OneShotModuleBufferize.h b/mlir/include/mlir/Dialect/Bufferization/Transforms/OneShotModuleBufferize.h index cd4c009..4e5f5e9 100644 --- a/mlir/include/mlir/Dialect/Bufferization/Transforms/OneShotModuleBufferize.h +++ b/mlir/include/mlir/Dialect/Bufferization/Transforms/OneShotModuleBufferize.h @@ -9,9 +9,11 @@ #ifndef MLIR_DIALECT_BUFFERIZATION_TRANSFORMS_ONESHOTMODULEBUFFERIZE_H #define MLIR_DIALECT_BUFFERIZATION_TRANSFORMS_ONESHOTMODULEBUFFERIZE_H -namespace mlir { - +namespace llvm { struct LogicalResult; +} // namespace llvm + +namespace mlir { class ModuleOp; namespace bufferization { @@ -21,8 +23,9 @@ struct OneShotBufferizationOptions; /// Analyze `moduleOp` and its nested ops. Bufferization decisions are stored in /// `state`. -LogicalResult analyzeModuleOp(ModuleOp moduleOp, OneShotAnalysisState &state, - BufferizationStatistics *statistics = nullptr); +llvm::LogicalResult +analyzeModuleOp(ModuleOp moduleOp, OneShotAnalysisState &state, + BufferizationStatistics *statistics = nullptr); /// Bufferize `op` and its nested ops that implement `BufferizableOpInterface`. /// @@ -33,9 +36,9 @@ LogicalResult analyzeModuleOp(ModuleOp moduleOp, OneShotAnalysisState &state, /// - `options.copyBeforeWrite` is not set and `options.noAnalysisFuncFilter` /// is not empty. The FuncOps it contains were not analyzed. Buffer copies /// will be inserted only to these FuncOps. -LogicalResult bufferizeModuleOp(ModuleOp moduleOp, - const OneShotBufferizationOptions &options, - BufferizationStatistics *statistics = nullptr); +llvm::LogicalResult +bufferizeModuleOp(ModuleOp moduleOp, const OneShotBufferizationOptions &options, + BufferizationStatistics *statistics = nullptr); /// Remove bufferization attributes on every FuncOp arguments in the ModuleOp. void removeBufferizationAttributesInModule(ModuleOp moduleOp); @@ -44,7 +47,7 @@ void removeBufferizationAttributesInModule(ModuleOp moduleOp); /// function call analysis to determine which function arguments are /// inplaceable. Then analyzes and bufferizes FuncOps one-by-one with One-Shot /// Bufferize. -LogicalResult runOneShotModuleBufferize( +llvm::LogicalResult runOneShotModuleBufferize( ModuleOp moduleOp, const bufferization::OneShotBufferizationOptions &options, BufferizationStatistics *statistics = nullptr); diff --git a/mlir/include/mlir/Dialect/GPU/IR/CompilationAttrInterfaces.td b/mlir/include/mlir/Dialect/GPU/IR/CompilationAttrInterfaces.td index 582ab91..43ca533 100644 --- a/mlir/include/mlir/Dialect/GPU/IR/CompilationAttrInterfaces.td +++ b/mlir/include/mlir/Dialect/GPU/IR/CompilationAttrInterfaces.td @@ -113,7 +113,7 @@ def OffloadingLLVMTranslationAttrInterface : The first argument has to be a GPU binary operation. If the function fails at any point, it must return `failure`. }], - "::mlir::LogicalResult", "embedBinary", + "::llvm::LogicalResult", "embedBinary", (ins "::mlir::Operation*":$binaryOp, "::llvm::IRBuilderBase&":$hostBuilder, "::mlir::LLVM::ModuleTranslation&":$hostModuleTranslation) @@ -130,7 +130,7 @@ def OffloadingLLVMTranslationAttrInterface : respectively. If the function fails at any point, it must return `failure`. }], - "::mlir::LogicalResult", "launchKernel", + "::llvm::LogicalResult", "launchKernel", (ins "::mlir::Operation*":$launchFunc, "::mlir::Operation*":$binaryOp, "::llvm::IRBuilderBase&":$hostBuilder, "::mlir::LLVM::ModuleTranslation&":$hostModuleTranslation) diff --git a/mlir/include/mlir/Dialect/GPU/Transforms/ParallelLoopMapper.h b/mlir/include/mlir/Dialect/GPU/Transforms/ParallelLoopMapper.h index 74b78f9..30511e7 100644 --- a/mlir/include/mlir/Dialect/GPU/Transforms/ParallelLoopMapper.h +++ b/mlir/include/mlir/Dialect/GPU/Transforms/ParallelLoopMapper.h @@ -21,7 +21,6 @@ namespace mlir { class AffineMap; -struct LogicalResult; class Operation; class Region; diff --git a/mlir/include/mlir/Dialect/GPU/Transforms/Utils.h b/mlir/include/mlir/Dialect/GPU/Transforms/Utils.h index f8c018e..0734939 100644 --- a/mlir/include/mlir/Dialect/GPU/Transforms/Utils.h +++ b/mlir/include/mlir/Dialect/GPU/Transforms/Utils.h @@ -20,7 +20,6 @@ #include <string> namespace mlir { -struct LogicalResult; class Operation; class Value; diff --git a/mlir/include/mlir/Dialect/IRDL/IR/IRDLInterfaces.h b/mlir/include/mlir/Dialect/IRDL/IR/IRDLInterfaces.h index 6455385..554fccec 100644 --- a/mlir/include/mlir/Dialect/IRDL/IR/IRDLInterfaces.h +++ b/mlir/include/mlir/Dialect/IRDL/IR/IRDLInterfaces.h @@ -19,7 +19,6 @@ #include "mlir/IR/ExtensibleDialect.h" #include "mlir/IR/OpImplementation.h" #include "mlir/IR/Types.h" -#include "mlir/Support/LogicalResult.h" #include <optional> namespace mlir { diff --git a/mlir/include/mlir/Dialect/IRDL/IR/IRDLTraits.h b/mlir/include/mlir/Dialect/IRDL/IR/IRDLTraits.h index 1755341..98bedbf 100644 --- a/mlir/include/mlir/Dialect/IRDL/IR/IRDLTraits.h +++ b/mlir/include/mlir/Dialect/IRDL/IR/IRDLTraits.h @@ -15,7 +15,6 @@ #define MLIR_DIALECT_IRDL_IR_IRDLTRAITS_H_ #include "mlir/IR/OpDefinition.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/Casting.h" namespace mlir { diff --git a/mlir/include/mlir/Dialect/IRDL/IRDLLoading.h b/mlir/include/mlir/Dialect/IRDL/IRDLLoading.h index 64ad72b..13c2799 100644 --- a/mlir/include/mlir/Dialect/IRDL/IRDLLoading.h +++ b/mlir/include/mlir/Dialect/IRDL/IRDLLoading.h @@ -13,8 +13,11 @@ #ifndef MLIR_DIALECT_IRDL_IRDLREGISTRATION_H #define MLIR_DIALECT_IRDL_IRDLREGISTRATION_H -namespace mlir { +namespace llvm { struct LogicalResult; +} // namespace llvm + +namespace mlir { class ModuleOp; } // namespace mlir @@ -22,7 +25,7 @@ namespace mlir { namespace irdl { /// Load all the dialects defined in the module. -LogicalResult loadDialects(ModuleOp op); +llvm::LogicalResult loadDialects(ModuleOp op); } // namespace irdl } // namespace mlir diff --git a/mlir/include/mlir/Dialect/IRDL/IRDLVerifiers.h b/mlir/include/mlir/Dialect/IRDL/IRDLVerifiers.h index 89e99a6..bad52bb 100644 --- a/mlir/include/mlir/Dialect/IRDL/IRDLVerifiers.h +++ b/mlir/include/mlir/Dialect/IRDL/IRDLVerifiers.h @@ -21,7 +21,6 @@ #include <optional> namespace mlir { -struct LogicalResult; class InFlightDiagnostic; class DynamicAttrDefinition; class DynamicTypeDefinition; diff --git a/mlir/include/mlir/Dialect/Linalg/TransformOps/GPUHeuristics.h b/mlir/include/mlir/Dialect/Linalg/TransformOps/GPUHeuristics.h index f9fd32c..5430fd9 100644 --- a/mlir/include/mlir/Dialect/Linalg/TransformOps/GPUHeuristics.h +++ b/mlir/include/mlir/Dialect/Linalg/TransformOps/GPUHeuristics.h @@ -11,7 +11,6 @@ #include "mlir/IR/Attributes.h" #include "mlir/IR/MLIRContext.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { namespace transform { diff --git a/mlir/include/mlir/Dialect/Linalg/TransformOps/Syntax.h b/mlir/include/mlir/Dialect/Linalg/TransformOps/Syntax.h index 13b0dc0..50e55e7 100644 --- a/mlir/include/mlir/Dialect/Linalg/TransformOps/Syntax.h +++ b/mlir/include/mlir/Dialect/Linalg/TransformOps/Syntax.h @@ -12,7 +12,6 @@ #include "mlir/Support/LLVM.h" namespace mlir { -class ParseResult; class OpAsmParser; class OpAsmPrinter; class Type; diff --git a/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h b/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h index 3812eb5..2a58d02 100644 --- a/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h +++ b/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h @@ -22,7 +22,6 @@ #include "mlir/Dialect/X86Vector/Transforms.h" #include "mlir/IR/PatternMatch.h" #include "mlir/Interfaces/TilingInterface.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/ADT/SmallBitVector.h" #include "llvm/ADT/SmallSet.h" diff --git a/mlir/include/mlir/Dialect/MemRef/Transforms/Transforms.h b/mlir/include/mlir/Dialect/MemRef/Transforms/Transforms.h index a918f62..4444a76 100644 --- a/mlir/include/mlir/Dialect/MemRef/Transforms/Transforms.h +++ b/mlir/include/mlir/Dialect/MemRef/Transforms/Transforms.h @@ -14,7 +14,7 @@ #ifndef MLIR_DIALECT_MEMREF_TRANSFORMS_TRANSFORMS_H #define MLIR_DIALECT_MEMREF_TRANSFORMS_TRANSFORMS_H -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "llvm/ADT/STLFunctionalExtras.h" namespace mlir { diff --git a/mlir/include/mlir/Dialect/Mesh/Transforms/Spmdization.h b/mlir/include/mlir/Dialect/Mesh/Transforms/Spmdization.h index f847ce3..2f6de3e 100644 --- a/mlir/include/mlir/Dialect/Mesh/Transforms/Spmdization.h +++ b/mlir/include/mlir/Dialect/Mesh/Transforms/Spmdization.h @@ -11,7 +11,6 @@ #include "mlir/Dialect/Mesh/IR/MeshOps.h" #include "mlir/IR/DialectRegistry.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { namespace mesh { diff --git a/mlir/include/mlir/Dialect/NVGPU/Transforms/Transforms.h b/mlir/include/mlir/Dialect/NVGPU/Transforms/Transforms.h index 4bac898..2ff9ac4 100644 --- a/mlir/include/mlir/Dialect/NVGPU/Transforms/Transforms.h +++ b/mlir/include/mlir/Dialect/NVGPU/Transforms/Transforms.h @@ -14,7 +14,6 @@ #define MLIR_DIALECT_NVGPU_TRANSFORMS_TRANSFORMS_H_ #include "mlir/IR/Operation.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { class RewriterBase; @@ -44,7 +43,7 @@ namespace nvgpu { /// function that depends on the row Index. The permutation function is chosen /// to ensure that sequential distributed+vectorized reads/writes down a single /// dimension of the memref have minimal conflicts. -mlir::LogicalResult optimizeSharedMemoryReadsAndWrites(Operation *parentOp, +llvm::LogicalResult optimizeSharedMemoryReadsAndWrites(Operation *parentOp, Value memrefValue); /// diff --git a/mlir/include/mlir/Dialect/OpenACCMPCommon/Interfaces/AtomicInterfaces.td b/mlir/include/mlir/Dialect/OpenACCMPCommon/Interfaces/AtomicInterfaces.td index 57063ca..223bee9 100644 --- a/mlir/include/mlir/Dialect/OpenACCMPCommon/Interfaces/AtomicInterfaces.td +++ b/mlir/include/mlir/Dialect/OpenACCMPCommon/Interfaces/AtomicInterfaces.td @@ -6,7 +6,7 @@ // //===----------------------------------------------------------------------===// // -// Defines the operation interface for atomic operations used in OpenACC and +// Defines the operation interface for atomic operations used in OpenACC and // OpenMP. // //===----------------------------------------------------------------------===// @@ -34,7 +34,7 @@ def AtomicReadOpInterface : OpInterface<"AtomicReadOpInterface"> { InterfaceMethod<[{ Common verifier for operation that implements atomic read interface. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyCommon", /*args=*/(ins), /*methodBody=*/"", @@ -81,7 +81,7 @@ def AtomicWriteOpInterface : OpInterface<"AtomicWriteOpInterface"> { InterfaceMethod<[{ Common verifier for operation that implements atomic write interface. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyCommon", /*args=*/(ins), /*methodBody=*/"", @@ -118,7 +118,7 @@ def AtomicUpdateOpInterface : OpInterface<"AtomicUpdateOpInterface"> { The interface terminology uses `x` to specify the address where a value is atomically written/read. - + Since atomic update expression comes in many forms, this interface requires that the operation uses a region with a single argument to capture the expression. @@ -162,9 +162,9 @@ def AtomicUpdateOpInterface : OpInterface<"AtomicUpdateOpInterface"> { /*defaultImplementation=*/[{ // The atomic update is a no-op if the terminator is the first and only // operation in its region. - mlir::Operation* terminator = + mlir::Operation* terminator = llvm::dyn_cast<mlir::RegionBranchTerminatorOpInterface>($_op.getFirstOp()); - return terminator && terminator->getOperands().front() == + return terminator && terminator->getOperands().front() == $_op.getRegion().front().getArgument(0); }] >, @@ -177,9 +177,9 @@ def AtomicUpdateOpInterface : OpInterface<"AtomicUpdateOpInterface"> { /*args=*/(ins), /*methodBody=*/"", /*defaultImplementation=*/[{ - mlir::Operation* terminator = + mlir::Operation* terminator = llvm::dyn_cast<mlir::RegionBranchTerminatorOpInterface>($_op.getFirstOp()); - if (terminator && terminator->getOperands().front() != + if (terminator && terminator->getOperands().front() != $_op.getRegion().front().getArgument(0)) { return terminator->getOperands().front(); } @@ -189,7 +189,7 @@ def AtomicUpdateOpInterface : OpInterface<"AtomicUpdateOpInterface"> { InterfaceMethod<[{ Common verifier for operation that implements atomic update interface. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyCommon", /*args=*/(ins), /*methodBody=*/"", @@ -210,7 +210,7 @@ def AtomicUpdateOpInterface : OpInterface<"AtomicUpdateOpInterface"> { Common verifier of the required region for operation that implements atomic update interface. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyRegionsCommon", /*args=*/(ins), /*methodBody=*/"", @@ -219,7 +219,7 @@ def AtomicUpdateOpInterface : OpInterface<"AtomicUpdateOpInterface"> { if (terminator->getOperands().size() != 1) return $_op.emitError("only updated value must be returned"); - + if (terminator->getOperands().front().getType() != $_op.getRegion().getArgument(0).getType()) return $_op.emitError("input and yielded value must have the same type"); @@ -272,7 +272,7 @@ def AtomicCaptureOpInterface : OpInterface<"AtomicCaptureOpInterface"> { Common verifier of the required region for operation that implements atomic capture interface. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyRegionsCommon", /*args=*/(ins), /*methodBody=*/"", diff --git a/mlir/include/mlir/Dialect/Polynomial/IR/Polynomial.h b/mlir/include/mlir/Dialect/Polynomial/IR/Polynomial.h index e14cef5..3f206cd 100644 --- a/mlir/include/mlir/Dialect/Polynomial/IR/Polynomial.h +++ b/mlir/include/mlir/Dialect/Polynomial/IR/Polynomial.h @@ -10,7 +10,6 @@ #define MLIR_DIALECT_POLYNOMIAL_IR_POLYNOMIAL_H_ #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APFloat.h" #include "llvm/ADT/APInt.h" #include "llvm/ADT/ArrayRef.h" diff --git a/mlir/include/mlir/Dialect/SCF/Transforms/Transforms.h b/mlir/include/mlir/Dialect/SCF/Transforms/Transforms.h index 1863317..71835cd 100644 --- a/mlir/include/mlir/Dialect/SCF/Transforms/Transforms.h +++ b/mlir/include/mlir/Dialect/SCF/Transforms/Transforms.h @@ -18,8 +18,6 @@ #include "llvm/ADT/ArrayRef.h" namespace mlir { - -struct LogicalResult; class Region; class RewriterBase; class Operation; diff --git a/mlir/include/mlir/Dialect/SCF/Utils/AffineCanonicalizationUtils.h b/mlir/include/mlir/Dialect/SCF/Utils/AffineCanonicalizationUtils.h index 5022cbf..2913a74 100644 --- a/mlir/include/mlir/Dialect/SCF/Utils/AffineCanonicalizationUtils.h +++ b/mlir/include/mlir/Dialect/SCF/Utils/AffineCanonicalizationUtils.h @@ -15,11 +15,9 @@ #define MLIR_DIALECT_SCF_UTILS_AFFINECANONICALIZATIONUTILS_H_ #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" namespace mlir { class AffineMap; -struct LogicalResult; class Operation; class OpFoldResult; class RewriterBase; diff --git a/mlir/include/mlir/Dialect/SCF/Utils/Utils.h b/mlir/include/mlir/Dialect/SCF/Utils/Utils.h index fea151b..de807c3 100644 --- a/mlir/include/mlir/Dialect/SCF/Utils/Utils.h +++ b/mlir/include/mlir/Dialect/SCF/Utils/Utils.h @@ -16,7 +16,6 @@ #include "mlir/Dialect/SCF/IR/SCF.h" #include "mlir/IR/PatternMatch.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include <optional> diff --git a/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorInterfaces.td b/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorInterfaces.td index 05eed04..5c07e81 100644 --- a/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorInterfaces.td +++ b/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorInterfaces.td @@ -32,7 +32,7 @@ def StageWithSortSparseOpInterface : OpInterface<"StageWithSortSparseOp"> { /*methodBody=*/"">, InterfaceMethod< /*desc=*/"Stage the operation, return the final result value after staging.", - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"stageWithSort", /*args=*/(ins "::mlir::PatternRewriter &":$rewriter, "Value &":$tmpBuf), diff --git a/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.h b/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.h index b835304..842e244 100644 --- a/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.h +++ b/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.h @@ -14,7 +14,6 @@ #include "mlir/IR/OpDefinition.h" #include "mlir/IR/PatternMatch.h" #include "mlir/Interfaces/SideEffectInterfaces.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "mlir/Dialect/Transform/Interfaces/TransformTypeInterfaces.h.inc" @@ -1595,7 +1594,7 @@ mlir::transform::TransformEachOpTrait<OpTy>::apply( } template <typename OpTy> -mlir::LogicalResult +llvm::LogicalResult mlir::transform::TransformEachOpTrait<OpTy>::verifyTrait(Operation *op) { static_assert(OpTy::template hasTrait<OpTrait::OneOperand>(), "expected single-operand op"); diff --git a/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.td b/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.td index c5c4c61..9439104 100644 --- a/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.td +++ b/mlir/include/mlir/Dialect/Transform/Interfaces/TransformInterfaces.td @@ -401,7 +401,7 @@ def ConversionPatternDescriptorOpInterface Verify the default type converter that is provided by the enclosing "apply_conversion_patterns" op. }], - /*returnType=*/"::mlir::LogicalResult", + /*returnType=*/"::llvm::LogicalResult", /*name=*/"verifyTypeConverter", /*arguments=*/(ins "TypeConverterBuilderOpInterface":$builder), /*methodBody=*/"", diff --git a/mlir/include/mlir/Dialect/Transform/Transforms/TransformInterpreterUtils.h b/mlir/include/mlir/Dialect/Transform/Transforms/TransformInterpreterUtils.h index 4c16d40..e78cf77 100644 --- a/mlir/include/mlir/Dialect/Transform/Transforms/TransformInterpreterUtils.h +++ b/mlir/include/mlir/Dialect/Transform/Transforms/TransformInterpreterUtils.h @@ -16,7 +16,6 @@ #include <memory> namespace mlir { -struct LogicalResult; class MLIRContext; class ModuleOp; class Operation; diff --git a/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h b/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h index fa2912a..8e6d36f 100644 --- a/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h +++ b/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h @@ -15,7 +15,6 @@ #include "mlir/Dialect/Vector/IR/VectorOps.h" #include "mlir/Dialect/Vector/Utils/VectorUtils.h" #include "mlir/IR/PatternMatch.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Dialect/Vector/Transforms/VectorTransformsEnums.h.inc" diff --git a/mlir/include/mlir/ExecutionEngine/JitRunner.h b/mlir/include/mlir/ExecutionEngine/JitRunner.h index 2dde094..b4210f6 100644 --- a/mlir/include/mlir/ExecutionEngine/JitRunner.h +++ b/mlir/include/mlir/ExecutionEngine/JitRunner.h @@ -24,6 +24,7 @@ namespace llvm { class Module; class LLVMContext; +struct LogicalResult; namespace orc { class MangleAndInterner; @@ -34,7 +35,6 @@ namespace mlir { class DialectRegistry; class Operation; -struct LogicalResult; /// JitRunner command line options used by JitRunnerConfig methods struct JitRunnerOptions { @@ -48,8 +48,8 @@ struct JitRunnerOptions { struct JitRunnerConfig { /// MLIR transformer applied after parsing the input into MLIR IR and before /// passing the MLIR IR to the ExecutionEngine. - llvm::function_ref<LogicalResult(mlir::Operation *, - JitRunnerOptions &options)> + llvm::function_ref<llvm::LogicalResult(mlir::Operation *, + JitRunnerOptions &options)> mlirTransformer = nullptr; /// A custom function that is passed to ExecutionEngine. It processes MLIR and diff --git a/mlir/include/mlir/IR/Action.h b/mlir/include/mlir/IR/Action.h index 9359324d..2fad477 100644 --- a/mlir/include/mlir/IR/Action.h +++ b/mlir/include/mlir/IR/Action.h @@ -16,7 +16,6 @@ #define MLIR_IR_ACTION_H #include "mlir/IR/Unit.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/Sequence.h" diff --git a/mlir/include/mlir/IR/AffineExpr.h b/mlir/include/mlir/IR/AffineExpr.h index 63314cc..a93e74b 100644 --- a/mlir/include/mlir/IR/AffineExpr.h +++ b/mlir/include/mlir/IR/AffineExpr.h @@ -20,7 +20,6 @@ #include "llvm/ADT/Hashing.h" #include "llvm/ADT/SmallVector.h" #include "llvm/Support/Casting.h" -#include <functional> #include <type_traits> namespace mlir { diff --git a/mlir/include/mlir/IR/AffineExprVisitor.h b/mlir/include/mlir/IR/AffineExprVisitor.h index fc4cd915..1826f5f 100644 --- a/mlir/include/mlir/IR/AffineExprVisitor.h +++ b/mlir/include/mlir/IR/AffineExprVisitor.h @@ -14,7 +14,7 @@ #define MLIR_IR_AFFINEEXPRVISITOR_H #include "mlir/IR/AffineExpr.h" -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "llvm/ADT/ArrayRef.h" namespace mlir { diff --git a/mlir/include/mlir/IR/AffineMap.h b/mlir/include/mlir/IR/AffineMap.h index cce1412..264c1c8 100644 --- a/mlir/include/mlir/IR/AffineMap.h +++ b/mlir/include/mlir/IR/AffineMap.h @@ -35,7 +35,6 @@ struct AffineMapStorage; class Attribute; class Builder; -struct LogicalResult; class OpFoldResult; class MLIRContext; diff --git a/mlir/include/mlir/IR/BuiltinAttributeInterfaces.h b/mlir/include/mlir/IR/BuiltinAttributeInterfaces.h index b0b9493..c4a4202 100644 --- a/mlir/include/mlir/IR/BuiltinAttributeInterfaces.h +++ b/mlir/include/mlir/IR/BuiltinAttributeInterfaces.h @@ -13,7 +13,6 @@ #include "mlir/IR/Attributes.h" #include "mlir/IR/BuiltinTypeInterfaces.h" #include "mlir/IR/Types.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/raw_ostream.h" #include <complex> #include <optional> diff --git a/mlir/include/mlir/IR/BuiltinAttributeInterfaces.td b/mlir/include/mlir/IR/BuiltinAttributeInterfaces.td index c741db9..954429c 100644 --- a/mlir/include/mlir/IR/BuiltinAttributeInterfaces.td +++ b/mlir/include/mlir/IR/BuiltinAttributeInterfaces.td @@ -483,7 +483,7 @@ def MemRefLayoutAttrInterface : AttrInterface<"MemRefLayoutAttrInterface"> { InterfaceMethod< "Check if the current layout is applicable to the provided shape", - "::mlir::LogicalResult", "verifyLayout", + "::llvm::LogicalResult", "verifyLayout", (ins "::llvm::ArrayRef<int64_t>":$shape, "::llvm::function_ref<::mlir::InFlightDiagnostic()>":$emitError), [{}], diff --git a/mlir/include/mlir/IR/Diagnostics.h b/mlir/include/mlir/IR/Diagnostics.h index 205c5d9..bb2e1bb 100644 --- a/mlir/include/mlir/IR/Diagnostics.h +++ b/mlir/include/mlir/IR/Diagnostics.h @@ -25,7 +25,6 @@ class SourceMgr; namespace mlir { class DiagnosticEngine; -struct LogicalResult; class MLIRContext; class Operation; class OperationName; diff --git a/mlir/include/mlir/IR/Dialect.h b/mlir/include/mlir/IR/Dialect.h index f7c1f4d..f3e5f6d 100644 --- a/mlir/include/mlir/IR/Dialect.h +++ b/mlir/include/mlir/IR/Dialect.h @@ -17,9 +17,6 @@ #include "mlir/IR/OperationSupport.h" #include "mlir/Support/TypeID.h" -#include <map> -#include <tuple> - namespace mlir { class DialectAsmParser; class DialectAsmPrinter; diff --git a/mlir/include/mlir/IR/EnumAttr.td b/mlir/include/mlir/IR/EnumAttr.td index f4dc480..9fec28f 100644 --- a/mlir/include/mlir/IR/EnumAttr.td +++ b/mlir/include/mlir/IR/EnumAttr.td @@ -227,7 +227,7 @@ class IntEnumAttr<I intType, string name, string summary, stringToSymbolFnName # [{(enumKeyword); if (maybeEnum) return *maybeEnum; - return {(::mlir::LogicalResult)($_parser.emitError(loc) << "expected " }] # + return {(::llvm::LogicalResult)($_parser.emitError(loc) << "expected " }] # [{<< "}] # cppType # [{" << " to be one of: " << }] # !interleave(!foreach(enum, enumerants, "\"" # enum.str # "\""), [{ << ", " << }]) # [{)}; @@ -300,7 +300,7 @@ class BitEnumAttr<I intType, string name, string summary, auto maybeEnum = }] # cppNamespace # "::" # stringToSymbolFnName # [{(enumKeyword); if (!maybeEnum) { - return {(::mlir::LogicalResult)($_parser.emitError(loc) << }] # + return {(::llvm::LogicalResult)($_parser.emitError(loc) << }] # [{"expected " << "}] # cppType # [{" << " to be one of: " << }] # !interleave(!foreach(enum, enumerants, "\"" # enum.str # "\""), [{ << ", " << }]) # [{)}; diff --git a/mlir/include/mlir/IR/ExtensibleDialect.h b/mlir/include/mlir/IR/ExtensibleDialect.h index 5ae1e9a..494f3df 100644 --- a/mlir/include/mlir/IR/ExtensibleDialect.h +++ b/mlir/include/mlir/IR/ExtensibleDialect.h @@ -38,7 +38,6 @@ class DynamicType; class ExtensibleDialect; class MLIRContext; class OptionalParseResult; -class ParseResult; namespace detail { struct DynamicAttrStorage; diff --git a/mlir/include/mlir/IR/OperationSupport.h b/mlir/include/mlir/IR/OperationSupport.h index f8ab533..1b93f3d 100644 --- a/mlir/include/mlir/IR/OperationSupport.h +++ b/mlir/include/mlir/IR/OperationSupport.h @@ -51,7 +51,6 @@ class OpAsmPrinter; class OperandRange; class OperandRangeRange; class OpFoldResult; -class ParseResult; class Pattern; class Region; class ResultRange; diff --git a/mlir/include/mlir/IR/StorageUniquerSupport.h b/mlir/include/mlir/IR/StorageUniquerSupport.h index 982d522..fb64f15 100644 --- a/mlir/include/mlir/IR/StorageUniquerSupport.h +++ b/mlir/include/mlir/IR/StorageUniquerSupport.h @@ -16,7 +16,6 @@ #include "mlir/IR/AttrTypeSubElements.h" #include "mlir/IR/DialectRegistry.h" #include "mlir/Support/InterfaceSupport.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/StorageUniquer.h" #include "mlir/Support/TypeID.h" #include "llvm/ADT/FunctionExtras.h" diff --git a/mlir/include/mlir/IR/SymbolInterfaces.td b/mlir/include/mlir/IR/SymbolInterfaces.td index 60b3818..a8b04d0 100644 --- a/mlir/include/mlir/IR/SymbolInterfaces.td +++ b/mlir/include/mlir/IR/SymbolInterfaces.td @@ -121,7 +121,7 @@ def Symbol : OpInterface<"SymbolOpInterface"> { symbol 'newSymbol' that are nested within the given operation 'from'. Note: See mlir::SymbolTable::replaceAllSymbolUses for more details. }], - "::mlir::LogicalResult", "replaceAllSymbolUses", + "::llvm::LogicalResult", "replaceAllSymbolUses", (ins "::mlir::StringAttr":$newSymbol, "::mlir::Operation *":$from), [{}], /*defaultImplementation=*/[{ return ::mlir::SymbolTable::replaceAllSymbolUses(this->getOperation(), @@ -214,7 +214,7 @@ def SymbolUserOpInterface : OpInterface<"SymbolUserOpInterface"> { let methods = [ InterfaceMethod<"Verify the symbol uses held by this operation.", - "::mlir::LogicalResult", "verifySymbolUses", + "::llvm::LogicalResult", "verifySymbolUses", (ins "::mlir::SymbolTableCollection &":$symbolTable) >, ]; diff --git a/mlir/include/mlir/IR/TensorEncoding.td b/mlir/include/mlir/IR/TensorEncoding.td index 4907dcb..d8ccd1f 100644 --- a/mlir/include/mlir/IR/TensorEncoding.td +++ b/mlir/include/mlir/IR/TensorEncoding.td @@ -31,7 +31,7 @@ def VerifiableTensorEncoding : AttrInterface<"VerifiableTensorEncoding"> { given shape and element type. Generates a diagnostic using the supplied callback on failure. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyEncoding", /*args=*/(ins "::mlir::ArrayRef<int64_t>":$shape, diff --git a/mlir/include/mlir/IR/Verifier.h b/mlir/include/mlir/IR/Verifier.h index 1fcc99e..eb96d11 100644 --- a/mlir/include/mlir/IR/Verifier.h +++ b/mlir/include/mlir/IR/Verifier.h @@ -9,8 +9,9 @@ #ifndef MLIR_IR_VERIFIER_H #define MLIR_IR_VERIFIER_H +#include "mlir/Support/LLVM.h" + namespace mlir { -struct LogicalResult; class Operation; /// Perform (potentially expensive) checks of invariants, used to detect diff --git a/mlir/include/mlir/IR/Visitors.h b/mlir/include/mlir/IR/Visitors.h index fe98774..15abf25 100644 --- a/mlir/include/mlir/IR/Visitors.h +++ b/mlir/include/mlir/IR/Visitors.h @@ -14,7 +14,6 @@ #define MLIR_IR_VISITORS_H #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" namespace mlir { diff --git a/mlir/include/mlir/InitAllPasses.h b/mlir/include/mlir/InitAllPasses.h index fedd773..1b9c1b1 100644 --- a/mlir/include/mlir/InitAllPasses.h +++ b/mlir/include/mlir/InitAllPasses.h @@ -98,9 +98,7 @@ inline void registerAllPasses() { bufferization::registerBufferizationPipelines(); sparse_tensor::registerSparseTensorPipelines(); tosa::registerTosaToLinalgPipelines(); -#if LLVM_HAS_NVPTX_TARGET gpu::registerGPUToNVVMPipeline(); -#endif } } // namespace mlir diff --git a/mlir/include/mlir/Interfaces/DataLayoutInterfaces.td b/mlir/include/mlir/Interfaces/DataLayoutInterfaces.td index d4b8999..bc5080c 100644 --- a/mlir/include/mlir/Interfaces/DataLayoutInterfaces.td +++ b/mlir/include/mlir/Interfaces/DataLayoutInterfaces.td @@ -52,7 +52,7 @@ def DataLayoutEntryInterface : AttrInterface<"DataLayoutEntryInterface"> { InterfaceMethod< /*description=*/"Checks that the entry is well-formed, reports errors " "at the provided location.", - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyEntry", /*args=*/(ins "::mlir::Location":$loc), /*methodBody=*/"", @@ -166,7 +166,7 @@ def DataLayoutSpecInterface : AttrInterface<"DataLayoutSpecInterface"> { InterfaceMethod< /*description=*/"Verifies the validity of the specification and reports " "any errors at the given location.", - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifySpec", /*args=*/(ins "::mlir::Location":$loc), /*methodBody=*/"", @@ -232,7 +232,7 @@ def TargetDeviceSpecInterface : AttrInterface<"TargetDeviceSpecInterface"> { InterfaceMethod< /*description=*/"Checks that the entry is well-formed, reports errors " "at the provided location.", - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyEntry", /*args=*/(ins "::mlir::Location":$loc), /*methodBody=*/"", @@ -275,7 +275,7 @@ def TargetSystemSpecInterface : AttrInterface<"TargetSystemSpecInterface"> { InterfaceMethod< /*description=*/"Verifies the validity of the specification and " "reports any errors at the given location.", - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifySpec", /*args=*/(ins "::mlir::Location":$loc), /*methodBody=*/"", @@ -574,7 +574,7 @@ def DataLayoutTypeInterface : TypeInterface<"DataLayoutTypeInterface"> { InterfaceMethod< /*description=*/"Verifies that the given list of entries is valid for " "this type.", - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"verifyEntries", /*args=*/(ins "::mlir::DataLayoutEntryListRef":$entries, "::mlir::Location":$loc), diff --git a/mlir/include/mlir/Interfaces/FoldInterfaces.h b/mlir/include/mlir/Interfaces/FoldInterfaces.h index 6278e31..b3d1161 100644 --- a/mlir/include/mlir/Interfaces/FoldInterfaces.h +++ b/mlir/include/mlir/Interfaces/FoldInterfaces.h @@ -9,7 +9,6 @@ #define MLIR_INTERFACES_FOLDINTERFACES_H_ #include "mlir/IR/DialectInterface.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/SmallVector.h" diff --git a/mlir/include/mlir/Interfaces/FunctionInterfaces.td b/mlir/include/mlir/Interfaces/FunctionInterfaces.td index 873853e..697f951 100644 --- a/mlir/include/mlir/Interfaces/FunctionInterfaces.td +++ b/mlir/include/mlir/Interfaces/FunctionInterfaces.td @@ -83,7 +83,7 @@ def FunctionOpInterface : OpInterface<"FunctionOpInterface", [ Note: The default implementation merely checks that if the entry block exists, it has the same number and type of arguments as the function type. }], - "::mlir::LogicalResult", "verifyBody", (ins), + "::llvm::LogicalResult", "verifyBody", (ins), /*methodBody=*/[{}], /*defaultImplementation=*/[{ if ($_op.isExternal()) return success(); @@ -114,7 +114,7 @@ def FunctionOpInterface : OpInterface<"FunctionOpInterface", [ Verify the type attribute of the function for derived op-specific invariants. }], - "::mlir::LogicalResult", "verifyType", (ins), + "::llvm::LogicalResult", "verifyType", (ins), /*methodBody=*/[{}], /*defaultImplementation=*/[{ return success(); }]>, diff --git a/mlir/include/mlir/Interfaces/InferTypeOpInterface.td b/mlir/include/mlir/Interfaces/InferTypeOpInterface.td index a009e21..92d4a99 100644 --- a/mlir/include/mlir/Interfaces/InferTypeOpInterface.td +++ b/mlir/include/mlir/Interfaces/InferTypeOpInterface.td @@ -38,7 +38,7 @@ def InferTypeOpInterface : OpInterface<"InferTypeOpInterface"> { called with valid arguments, e.g., operands are verified, or it may result in an undefined behavior. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"inferReturnTypes", /*args=*/(ins "::mlir::MLIRContext *":$context, "::std::optional<::mlir::Location>":$location, @@ -72,7 +72,7 @@ def InferTypeOpInterface : OpInterface<"InferTypeOpInterface"> { represent fully valid IR and are responsible for checking inputs for validity to the degree necessary to perform the return type inference. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"refineReturnTypes", /*args=*/(ins "::mlir::MLIRContext *":$context, "::std::optional<::mlir::Location>":$location, @@ -150,7 +150,7 @@ def InferShapedTypeOpInterface : OpInterface<"InferShapedTypeOpInterface"> { represent fully valid IR and are responsible for checking inputs for validity to the degree necessary to perform the return type inference. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"inferReturnTypeComponents", /*args=*/(ins "::mlir::MLIRContext*":$context, "::std::optional<::mlir::Location>":$location, @@ -180,7 +180,7 @@ def InferShapedTypeOpInterface : OpInterface<"InferShapedTypeOpInterface"> { operands to avoid calling `getOperand` directly inside the interface implementation. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"reifyReturnTypeShapes", /*args=*/(ins "::mlir::OpBuilder&":$builder, "::mlir::ValueRange":$operands, @@ -201,14 +201,14 @@ class InferTypeOpAdaptorBase<code additionalDecls = [{}]> : TraitList< /*name=*/"InferTypeOpAdaptor", /*traits=*/[], /*extraOpDeclaration=*/[{ - static ::mlir::LogicalResult + static ::llvm::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, std::optional<::mlir::Location> location, Adaptor adaptor, ::llvm::SmallVectorImpl<::mlir::Type> &inferredReturnTypes); }] # additionalDecls, /*extraOpDefinition=*/[{ - ::mlir::LogicalResult + ::llvm::LogicalResult $cppClass::inferReturnTypes(::mlir::MLIRContext *context, std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, @@ -239,14 +239,14 @@ class InferShapedTypeOpAdaptorBase<list<string> overridenMethods = []> : TraitLi /*name=*/"InferShapedTypeOpAdaptor", /*traits=*/[], /*extraOpDeclaration=*/[{ - static ::mlir::LogicalResult + static ::llvm::LogicalResult inferReturnTypeComponents(::mlir::MLIRContext *context, std::optional<::mlir::Location> location, Adaptor adaptor, ::llvm::SmallVectorImpl<::mlir::ShapedTypeComponents> &inferredReturnShapes); }], /*extraOpDefinition=*/[{ - ::mlir::LogicalResult + ::llvm::LogicalResult $cppClass::inferReturnTypeComponents(::mlir::MLIRContext *context, std::optional<::mlir::Location> location, ::mlir::ValueShapeRange operands, ::mlir::DictionaryAttr attributes, @@ -281,7 +281,7 @@ class InferTensorTypeBase<list<string> overridenMethods = []> : TraitList< /*traits=*/[], /*extraOpDeclaration=*/[{}], /*extraOpDefinition=*/[{ - ::mlir::LogicalResult + ::llvm::LogicalResult $cppClass::inferReturnTypes(::mlir::MLIRContext *context, std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, @@ -364,7 +364,7 @@ def ReifyRankedShapedTypeOpInterface : If the shape of a particular result cannot be computed it must be empty. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"reifyResultShapes", /*args=*/(ins "::mlir::OpBuilder &":$builder, "::mlir::ReifiedRankedShapedTypeDims &":$reifiedReturnShapes) diff --git a/mlir/include/mlir/Interfaces/LoopLikeInterface.td b/mlir/include/mlir/Interfaces/LoopLikeInterface.td index 7db624b..c6bffe3 100644 --- a/mlir/include/mlir/Interfaces/LoopLikeInterface.td +++ b/mlir/include/mlir/Interfaces/LoopLikeInterface.td @@ -84,7 +84,7 @@ def LoopLikeOpInterface : OpInterface<"LoopLikeOpInterface"> { have a single iteration. Returns "success" if the promotion was successful. }], - /*retTy=*/"::mlir::LogicalResult", + /*retTy=*/"::llvm::LogicalResult", /*methodName=*/"promoteIfSingleIteration", /*args=*/(ins "::mlir::RewriterBase &":$rewriter), /*methodBody=*/"", diff --git a/mlir/include/mlir/Interfaces/MemorySlotInterfaces.td b/mlir/include/mlir/Interfaces/MemorySlotInterfaces.td index 6f023f0..fbce2fa 100644 --- a/mlir/include/mlir/Interfaces/MemorySlotInterfaces.td +++ b/mlir/include/mlir/Interfaces/MemorySlotInterfaces.td @@ -75,7 +75,7 @@ def PromotableAllocationOpInterface "::std::optional<::mlir::PromotableAllocationOpInterface>", "handlePromotionComplete", (ins - "const ::mlir::MemorySlot &":$slot, + "const ::mlir::MemorySlot &":$slot, "::mlir::Value":$defaultValue, "::mlir::OpBuilder &":$builder) >, @@ -308,7 +308,7 @@ def DestructurableAllocationOpInterface This will only be called for slots declared by this operation. Must return a new destructurable allocation op if this hook creates - a new destructurable op, nullopt otherwise. + a new destructurable op, nullopt otherwise. }], "::std::optional<::mlir::DestructurableAllocationOpInterface>", "handleDestructuringComplete", @@ -338,7 +338,7 @@ def SafeMemorySlotAccessOpInterface No IR mutation is allowed in this method. }], - "::mlir::LogicalResult", + "::llvm::LogicalResult", "ensureOnlySafeAccesses", (ins "const ::mlir::MemorySlot &":$slot, "::mlir::SmallVectorImpl<::mlir::MemorySlot> &":$mustBeSafelyUsed, diff --git a/mlir/include/mlir/Interfaces/TilingInterface.td b/mlir/include/mlir/Interfaces/TilingInterface.td index ad0af0a..4868dad 100644 --- a/mlir/include/mlir/Interfaces/TilingInterface.td +++ b/mlir/include/mlir/Interfaces/TilingInterface.td @@ -26,7 +26,7 @@ def TilingInterface : OpInterface<"TilingInterface"> { the tiling algorithm (like `scf::tileUsingSCF`) can generate the inter-tile loop structure, and call into the methods of the interface to be able to tile any operation that implements the interface. - + This interface is also meant to help with "tile and fuse", i.e. the process of fusing a producer with a consumer by a) Tiling the consumer @@ -50,7 +50,7 @@ def TilingInterface : OpInterface<"TilingInterface"> { For an operation to be "tiled and fused" with its (already tiled) consumer, an operation has to implement the following additional method (see - description below): + description below): - `generateResultTileValue` - `getIterationDomainTileFromResultTile` @@ -146,7 +146,7 @@ def TilingInterface : OpInterface<"TilingInterface"> { Note: It is undefined behaviour if there is overlap between the tiles of the result generated by the tiled implementation. }], - /*retType=*/"::mlir::LogicalResult", + /*retType=*/"::llvm::LogicalResult", /*methodName=*/"getResultTilePosition", /*args=*/(ins "OpBuilder &":$b, @@ -289,7 +289,7 @@ def TilingInterface : OpInterface<"TilingInterface"> { implemented using `getIterationDomainTileFromOperandTile` + `getTiledImplementation` methods. }], - /*retType=*/"::mlir::LogicalResult", + /*retType=*/"::llvm::LogicalResult", /*methodName=*/"getIterationDomainTileFromOperandTile", /*args=*/(ins "OpBuilder &":$b, @@ -317,14 +317,14 @@ def TilingInterface : OpInterface<"TilingInterface"> { - `sizes` is the size of the slice of the producer result used by the consumer. If fusion of the producer with the consumer is not legal for the - result, or if this mapping cannot be computed, the implementation + result, or if this mapping cannot be computed, the implementation should return a failure. - - For most cases `generateResultTileValue` could be a implemented using - `getIterationDomainTileFromResultTile` + `getTiledImplementation` + + For most cases `generateResultTileValue` could be a implemented using + `getIterationDomainTileFromResultTile` + `getTiledImplementation` methods. }], - /*retType=*/"::mlir::LogicalResult", + /*retType=*/"::llvm::LogicalResult", /*methodName=*/"getIterationDomainTileFromResultTile", /*args=*/(ins "OpBuilder &":$b, @@ -349,7 +349,7 @@ def TilingInterface : OpInterface<"TilingInterface"> { transformations are done, this method can be used to lower to scalar code that can then be lowered to LLVM or SPIR-V dialects. }], - /*retType=*/"::mlir::LogicalResult", + /*retType=*/"::llvm::LogicalResult", /*methodName=*/"generateScalarImplementation", /*args=*/(ins "OpBuilder &":$b, diff --git a/mlir/include/mlir/Pass/Pass.h b/mlir/include/mlir/Pass/Pass.h index e71c49a..7725a3a 100644 --- a/mlir/include/mlir/Pass/Pass.h +++ b/mlir/include/mlir/Pass/Pass.h @@ -12,7 +12,6 @@ #include "mlir/IR/Action.h" #include "mlir/Pass/AnalysisManager.h" #include "mlir/Pass/PassRegistry.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/PointerIntPair.h" #include "llvm/ADT/Statistic.h" #include <optional> diff --git a/mlir/include/mlir/Pass/PassManager.h b/mlir/include/mlir/Pass/PassManager.h index b3e4275..d9bab43 100644 --- a/mlir/include/mlir/Pass/PassManager.h +++ b/mlir/include/mlir/Pass/PassManager.h @@ -11,7 +11,6 @@ #include "mlir/IR/Dialect.h" #include "mlir/IR/OperationSupport.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/Timing.h" #include "llvm/ADT/SmallVector.h" #include "llvm/ADT/iterator.h" diff --git a/mlir/include/mlir/Pass/PassOptions.h b/mlir/include/mlir/Pass/PassOptions.h index 3a5e322..b99d91f 100644 --- a/mlir/include/mlir/Pass/PassOptions.h +++ b/mlir/include/mlir/Pass/PassOptions.h @@ -15,7 +15,6 @@ #define MLIR_PASS_PASSOPTIONS_H_ #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/FunctionExtras.h" #include "llvm/ADT/StringRef.h" #include "llvm/Support/CommandLine.h" diff --git a/mlir/include/mlir/Query/Query.h b/mlir/include/mlir/Query/Query.h index 447fc7c..18f2172 100644 --- a/mlir/include/mlir/Query/Query.h +++ b/mlir/include/mlir/Query/Query.h @@ -10,7 +10,6 @@ #define MLIR_TOOLS_MLIRQUERY_QUERY_H #include "Matcher/VariantValue.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/IntrusiveRefCntPtr.h" #include "llvm/ADT/StringRef.h" #include "llvm/LineEditor/LineEditor.h" @@ -27,7 +26,7 @@ struct Query : llvm::RefCountedBase<Query> { virtual ~Query(); // Perform the query on qs and print output to os. - virtual mlir::LogicalResult run(llvm::raw_ostream &os, + virtual llvm::LogicalResult run(llvm::raw_ostream &os, QuerySession &qs) const = 0; llvm::StringRef remainingContent; @@ -45,7 +44,7 @@ complete(llvm::StringRef line, size_t pos, const QuerySession &qs); struct InvalidQuery : Query { InvalidQuery(const llvm::Twine &errStr) : Query(QueryKind::Invalid), errStr(errStr.str()) {} - mlir::LogicalResult run(llvm::raw_ostream &os, + llvm::LogicalResult run(llvm::raw_ostream &os, QuerySession &qs) const override; std::string errStr; @@ -58,7 +57,7 @@ struct InvalidQuery : Query { // No-op query (i.e. a blank line). struct NoOpQuery : Query { NoOpQuery() : Query(QueryKind::NoOp) {} - mlir::LogicalResult run(llvm::raw_ostream &os, + llvm::LogicalResult run(llvm::raw_ostream &os, QuerySession &qs) const override; static bool classof(const Query *query) { @@ -69,7 +68,7 @@ struct NoOpQuery : Query { // Query for "help". struct HelpQuery : Query { HelpQuery() : Query(QueryKind::Help) {} - mlir::LogicalResult run(llvm::raw_ostream &os, + llvm::LogicalResult run(llvm::raw_ostream &os, QuerySession &qs) const override; static bool classof(const Query *query) { @@ -80,7 +79,7 @@ struct HelpQuery : Query { // Query for "quit". struct QuitQuery : Query { QuitQuery() : Query(QueryKind::Quit) {} - mlir::LogicalResult run(llvm::raw_ostream &os, + llvm::LogicalResult run(llvm::raw_ostream &os, QuerySession &qs) const override; static bool classof(const Query *query) { @@ -92,7 +91,7 @@ struct QuitQuery : Query { struct MatchQuery : Query { MatchQuery(llvm::StringRef source, const matcher::DynMatcher &matcher) : Query(QueryKind::Match), matcher(matcher), source(source) {} - mlir::LogicalResult run(llvm::raw_ostream &os, + llvm::LogicalResult run(llvm::raw_ostream &os, QuerySession &qs) const override; const matcher::DynMatcher matcher; diff --git a/mlir/include/mlir/Reducer/ReductionNode.h b/mlir/include/mlir/Reducer/ReductionNode.h index 814c885..6ca4e13 100644 --- a/mlir/include/mlir/Reducer/ReductionNode.h +++ b/mlir/include/mlir/Reducer/ReductionNode.h @@ -22,7 +22,6 @@ #include "mlir/IR/OwningOpRef.h" #include "mlir/Reducer/Tester.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/Support/Allocator.h" #include "llvm/Support/ToolOutputFile.h" diff --git a/mlir/include/mlir/Support/LLVM.h b/mlir/include/mlir/Support/LLVM.h index 7baca03..020c0fb 100644 --- a/mlir/include/mlir/Support/LLVM.h +++ b/mlir/include/mlir/Support/LLVM.h @@ -21,7 +21,10 @@ // We include this header because it cannot be practically forward // declared, and are effectively language features. #include "llvm/Support/Casting.h" -#include <vector> + +// We include this header because large portions of mlir would have to include +// it anyway. +#include "llvm/Support/LogicalResult.h" // Workaround for clang-5 (PR41549) #if defined(__clang_major__) @@ -151,6 +154,15 @@ using llvm::iterator_range; using llvm::raw_ostream; using llvm::SMLoc; using llvm::SMRange; + +// LogicalResult. +using llvm::failed; +using llvm::failure; +using llvm::FailureOr; +using llvm::LogicalResult; +using llvm::ParseResult; +using llvm::succeeded; +using llvm::success; } // namespace mlir #endif // MLIR_SUPPORT_LLVM_H diff --git a/mlir/include/mlir/Support/LogicalResult.h b/mlir/include/mlir/Support/LogicalResult.h index 7ee7046..c7da8de 100644 --- a/mlir/include/mlir/Support/LogicalResult.h +++ b/mlir/include/mlir/Support/LogicalResult.h @@ -1,4 +1,4 @@ -//===- LogicalResult.h - Utilities for handling success/failure -*- C++ -*-===// +//===- LogicalResult.h - Stub aliasing to llvm/LogicalResult ----*- C++ -*-===// // // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. // See https://llvm.org/LICENSE.txt for license information. @@ -9,123 +9,18 @@ #ifndef MLIR_SUPPORT_LOGICALRESULT_H #define MLIR_SUPPORT_LOGICALRESULT_H -#include "mlir/Support/LLVM.h" -#include <optional> +#include "llvm/Support/LogicalResult.h" +// TODO: This header is a stop-gap to avoid breaking downstream, and is to be +// removed eventually. namespace mlir { - -/// This class represents an efficient way to signal success or failure. It -/// should be preferred over the use of `bool` when appropriate, as it avoids -/// all of the ambiguity that arises in interpreting a boolean result. This -/// class is marked as NODISCARD to ensure that the result is processed. Users -/// may explicitly discard a result by using `(void)`, e.g. -/// `(void)functionThatReturnsALogicalResult();`. Given the intended nature of -/// this class, it generally shouldn't be used as the result of functions that -/// very frequently have the result ignored. This class is intended to be used -/// in conjunction with the utility functions below. -struct [[nodiscard]] LogicalResult { -public: - /// If isSuccess is true a `success` result is generated, otherwise a - /// 'failure' result is generated. - static LogicalResult success(bool isSuccess = true) { - return LogicalResult(isSuccess); - } - - /// If isFailure is true a `failure` result is generated, otherwise a - /// 'success' result is generated. - static LogicalResult failure(bool isFailure = true) { - return LogicalResult(!isFailure); - } - - /// Returns true if the provided LogicalResult corresponds to a success value. - bool succeeded() const { return isSuccess; } - - /// Returns true if the provided LogicalResult corresponds to a failure value. - bool failed() const { return !isSuccess; } - -private: - LogicalResult(bool isSuccess) : isSuccess(isSuccess) {} - - /// Boolean indicating if this is a success result, if false this is a - /// failure result. - bool isSuccess; -}; - -/// Utility function to generate a LogicalResult. If isSuccess is true a -/// `success` result is generated, otherwise a 'failure' result is generated. -inline LogicalResult success(bool isSuccess = true) { - return LogicalResult::success(isSuccess); -} - -/// Utility function to generate a LogicalResult. If isFailure is true a -/// `failure` result is generated, otherwise a 'success' result is generated. -inline LogicalResult failure(bool isFailure = true) { - return LogicalResult::failure(isFailure); -} - -/// Utility function that returns true if the provided LogicalResult corresponds -/// to a success value. -inline bool succeeded(LogicalResult result) { return result.succeeded(); } - -/// Utility function that returns true if the provided LogicalResult corresponds -/// to a failure value. -inline bool failed(LogicalResult result) { return result.failed(); } - -/// This class provides support for representing a failure result, or a valid -/// value of type `T`. This allows for integrating with LogicalResult, while -/// also providing a value on the success path. -template <typename T> -class [[nodiscard]] FailureOr : public std::optional<T> { -public: - /// Allow constructing from a LogicalResult. The result *must* be a failure. - /// Success results should use a proper instance of type `T`. - FailureOr(LogicalResult result) { - assert(failed(result) && - "success should be constructed with an instance of 'T'"); - } - FailureOr() : FailureOr(failure()) {} - FailureOr(T &&y) : std::optional<T>(std::forward<T>(y)) {} - FailureOr(const T &y) : std::optional<T>(y) {} - template <typename U, - std::enable_if_t<std::is_constructible<T, U>::value> * = nullptr> - FailureOr(const FailureOr<U> &other) - : std::optional<T>(failed(other) ? std::optional<T>() - : std::optional<T>(*other)) {} - - operator LogicalResult() const { return success(this->has_value()); } - -private: - /// Hide the bool conversion as it easily creates confusion. - using std::optional<T>::operator bool; - using std::optional<T>::has_value; -}; - -/// Wrap a value on the success path in a FailureOr of the same value type. -template <typename T, - typename = std::enable_if_t<!std::is_convertible_v<T, bool>>> -inline auto success(T &&t) { - return FailureOr<std::decay_t<T>>(std::forward<T>(t)); -} - -/// This class represents success/failure for parsing-like operations that find -/// it important to chain together failable operations with `||`. This is an -/// extended version of `LogicalResult` that allows for explicit conversion to -/// bool. -/// -/// This class should not be used for general error handling cases - we prefer -/// to keep the logic explicit with the `succeeded`/`failed` predicates. -/// However, traditional monadic-style parsing logic can sometimes get -/// swallowed up in boilerplate without this, so we provide this for narrow -/// cases where it is important. -/// -class [[nodiscard]] ParseResult : public LogicalResult { -public: - ParseResult(LogicalResult result = success()) : LogicalResult(result) {} - - /// Failure is true in a boolean context. - explicit operator bool() const { return failed(); } -}; - +using llvm::failed; +using llvm::failure; +using llvm::FailureOr; +using llvm::LogicalResult; +using llvm::ParseResult; +using llvm::succeeded; +using llvm::success; } // namespace mlir #endif // MLIR_SUPPORT_LOGICALRESULT_H diff --git a/mlir/include/mlir/Support/StorageUniquer.h b/mlir/include/mlir/Support/StorageUniquer.h index 2369d9c..26bdf09 100644 --- a/mlir/include/mlir/Support/StorageUniquer.h +++ b/mlir/include/mlir/Support/StorageUniquer.h @@ -10,7 +10,6 @@ #define MLIR_SUPPORT_STORAGEUNIQUER_H #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/DenseSet.h" diff --git a/mlir/include/mlir/Support/ToolUtilities.h b/mlir/include/mlir/Support/ToolUtilities.h index 511cb11..cb6ba29 100644 --- a/mlir/include/mlir/Support/ToolUtilities.h +++ b/mlir/include/mlir/Support/ToolUtilities.h @@ -24,8 +24,6 @@ class MemoryBuffer; } // namespace llvm namespace mlir { -struct LogicalResult; - using ChunkBufferHandler = function_ref<LogicalResult( std::unique_ptr<llvm::MemoryBuffer> chunkBuffer, raw_ostream &os)>; diff --git a/mlir/include/mlir/Target/Cpp/CppEmitter.h b/mlir/include/mlir/Target/Cpp/CppEmitter.h index d3c0810..99d8696 100644 --- a/mlir/include/mlir/Target/Cpp/CppEmitter.h +++ b/mlir/include/mlir/Target/Cpp/CppEmitter.h @@ -16,7 +16,6 @@ #include "mlir/Support/LLVM.h" namespace mlir { -struct LogicalResult; class Operation; namespace emitc { diff --git a/mlir/include/mlir/Target/LLVM/ROCDL/Utils.h b/mlir/include/mlir/Target/LLVM/ROCDL/Utils.h index 44c9ded..3c637a0 100644 --- a/mlir/include/mlir/Target/LLVM/ROCDL/Utils.h +++ b/mlir/include/mlir/Target/LLVM/ROCDL/Utils.h @@ -15,6 +15,7 @@ #include "mlir/Dialect/GPU/IR/CompilationInterfaces.h" #include "mlir/Dialect/LLVMIR/ROCDLDialect.h" +#include "mlir/Support/LLVM.h" #include "mlir/Target/LLVM/ModuleToObject.h" namespace mlir { diff --git a/mlir/include/mlir/Target/LLVMIR/LLVMImportInterface.h b/mlir/include/mlir/Target/LLVMIR/LLVMImportInterface.h index 86bcd58..74b5454 100644 --- a/mlir/include/mlir/Target/LLVMIR/LLVMImportInterface.h +++ b/mlir/include/mlir/Target/LLVMIR/LLVMImportInterface.h @@ -19,7 +19,6 @@ #include "mlir/IR/Diagnostics.h" #include "mlir/IR/DialectInterface.h" #include "mlir/IR/Location.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/IR/Instruction.h" #include "llvm/IR/Instructions.h" #include "llvm/Support/FormatVariadic.h" diff --git a/mlir/include/mlir/Target/LLVMIR/LLVMTranslationInterface.h b/mlir/include/mlir/Target/LLVMIR/LLVMTranslationInterface.h index 4a8ee06..11cb836 100644 --- a/mlir/include/mlir/Target/LLVMIR/LLVMTranslationInterface.h +++ b/mlir/include/mlir/Target/LLVMIR/LLVMTranslationInterface.h @@ -16,7 +16,6 @@ #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/IR/BuiltinAttributes.h" #include "mlir/IR/DialectInterface.h" -#include "mlir/Support/LogicalResult.h" namespace llvm { class Instruction; diff --git a/mlir/include/mlir/Target/SPIRV/Serialization.h b/mlir/include/mlir/Target/SPIRV/Serialization.h index 498f390..613f0a4 100644 --- a/mlir/include/mlir/Target/SPIRV/Serialization.h +++ b/mlir/include/mlir/Target/SPIRV/Serialization.h @@ -16,7 +16,6 @@ #include "mlir/Support/LLVM.h" namespace mlir { -struct LogicalResult; class MLIRContext; namespace spirv { diff --git a/mlir/include/mlir/Tools/PDLL/AST/Diagnostic.h b/mlir/include/mlir/Tools/PDLL/AST/Diagnostic.h index a8bae32..aaba810 100644 --- a/mlir/include/mlir/Tools/PDLL/AST/Diagnostic.h +++ b/mlir/include/mlir/Tools/PDLL/AST/Diagnostic.h @@ -13,7 +13,6 @@ #include <optional> #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/FunctionExtras.h" #include "llvm/Support/SourceMgr.h" diff --git a/mlir/include/mlir/Tools/PDLL/CodeGen/MLIRGen.h b/mlir/include/mlir/Tools/PDLL/CodeGen/MLIRGen.h index c691c39..c106f6d 100644 --- a/mlir/include/mlir/Tools/PDLL/CodeGen/MLIRGen.h +++ b/mlir/include/mlir/Tools/PDLL/CodeGen/MLIRGen.h @@ -11,8 +11,6 @@ #include <memory> -#include "mlir/Support/LogicalResult.h" - namespace llvm { class SourceMgr; } // namespace llvm diff --git a/mlir/include/mlir/Tools/PDLL/Parser/Parser.h b/mlir/include/mlir/Tools/PDLL/Parser/Parser.h index 1a43a3b..6f62aaf 100644 --- a/mlir/include/mlir/Tools/PDLL/Parser/Parser.h +++ b/mlir/include/mlir/Tools/PDLL/Parser/Parser.h @@ -11,7 +11,7 @@ #include <memory> -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" namespace llvm { class SourceMgr; diff --git a/mlir/include/mlir/Tools/lsp-server-support/CompilationDatabase.h b/mlir/include/mlir/Tools/lsp-server-support/CompilationDatabase.h index 84807ec..588c136 100644 --- a/mlir/include/mlir/Tools/lsp-server-support/CompilationDatabase.h +++ b/mlir/include/mlir/Tools/lsp-server-support/CompilationDatabase.h @@ -20,6 +20,7 @@ #include "llvm/ADT/StringMap.h" #include <memory> #include <string> +#include <vector> namespace mlir { namespace lsp { diff --git a/mlir/include/mlir/Tools/lsp-server-support/Protocol.h b/mlir/include/mlir/Tools/lsp-server-support/Protocol.h index 1d22b8a..5d2eb01 100644 --- a/mlir/include/mlir/Tools/lsp-server-support/Protocol.h +++ b/mlir/include/mlir/Tools/lsp-server-support/Protocol.h @@ -28,15 +28,12 @@ #include "llvm/Support/SourceMgr.h" #include "llvm/Support/raw_ostream.h" #include <bitset> -#include <memory> #include <optional> #include <string> #include <utility> #include <vector> namespace mlir { -struct LogicalResult; - namespace lsp { enum class ErrorCode { diff --git a/mlir/include/mlir/Tools/lsp-server-support/Transport.h b/mlir/include/mlir/Tools/lsp-server-support/Transport.h index 83fff92..6843bc7 100644 --- a/mlir/include/mlir/Tools/lsp-server-support/Transport.h +++ b/mlir/include/mlir/Tools/lsp-server-support/Transport.h @@ -17,7 +17,6 @@ #include "mlir/Support/DebugStringHelper.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Tools/lsp-server-support/Logging.h" #include "mlir/Tools/lsp-server-support/Protocol.h" #include "llvm/ADT/FunctionExtras.h" diff --git a/mlir/include/mlir/Tools/mlir-lsp-server/MlirLspServerMain.h b/mlir/include/mlir/Tools/mlir-lsp-server/MlirLspServerMain.h index 5207106..66d5d40 100644 --- a/mlir/include/mlir/Tools/mlir-lsp-server/MlirLspServerMain.h +++ b/mlir/include/mlir/Tools/mlir-lsp-server/MlirLspServerMain.h @@ -13,15 +13,18 @@ #ifndef MLIR_TOOLS_MLIR_LSP_SERVER_MLIRLSPSERVERMAIN_H #define MLIR_TOOLS_MLIR_LSP_SERVER_MLIRLSPSERVERMAIN_H +namespace llvm { +struct LogicalResult; +} // namespace llvm + namespace mlir { class DialectRegistry; -struct LogicalResult; /// Implementation for tools like `mlir-lsp-server`. /// - registry should contain all the dialects that can be parsed in source IR /// passed to the server. -LogicalResult MlirLspServerMain(int argc, char **argv, - DialectRegistry ®istry); +llvm::LogicalResult MlirLspServerMain(int argc, char **argv, + DialectRegistry ®istry); } // namespace mlir diff --git a/mlir/include/mlir/Tools/mlir-opt/MlirOptMain.h b/mlir/include/mlir/Tools/mlir-opt/MlirOptMain.h index 4f7f83c..d0ca188 100644 --- a/mlir/include/mlir/Tools/mlir-opt/MlirOptMain.h +++ b/mlir/include/mlir/Tools/mlir-opt/MlirOptMain.h @@ -14,7 +14,6 @@ #define MLIR_TOOLS_MLIROPT_MLIROPTMAIN_H #include "mlir/Debug/CLOptionsSetup.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/ToolUtilities.h" #include "llvm/ADT/StringRef.h" diff --git a/mlir/include/mlir/Tools/mlir-pdll-lsp-server/MlirPdllLspServerMain.h b/mlir/include/mlir/Tools/mlir-pdll-lsp-server/MlirPdllLspServerMain.h index a193732..c45ddb6 100644 --- a/mlir/include/mlir/Tools/mlir-pdll-lsp-server/MlirPdllLspServerMain.h +++ b/mlir/include/mlir/Tools/mlir-pdll-lsp-server/MlirPdllLspServerMain.h @@ -14,11 +14,13 @@ #ifndef MLIR_TOOLS_MLIR_PDLL_LSP_SERVER_MLIRPDLLLSPSERVERMAIN_H #define MLIR_TOOLS_MLIR_PDLL_LSP_SERVER_MLIRPDLLLSPSERVERMAIN_H -namespace mlir { +namespace llvm { struct LogicalResult; +} // namespace llvm +namespace mlir { /// Implementation for tools like `mlir-pdll-lsp-server`. -LogicalResult MlirPdllLspServerMain(int argc, char **argv); +llvm::LogicalResult MlirPdllLspServerMain(int argc, char **argv); } // namespace mlir diff --git a/mlir/include/mlir/Tools/mlir-query/MlirQueryMain.h b/mlir/include/mlir/Tools/mlir-query/MlirQueryMain.h index fa1cd5d..44d5c8d 100644 --- a/mlir/include/mlir/Tools/mlir-query/MlirQueryMain.h +++ b/mlir/include/mlir/Tools/mlir-query/MlirQueryMain.h @@ -15,7 +15,7 @@ #define MLIR_TOOLS_MLIRQUERY_MLIRQUERYMAIN_H #include "mlir/Query/Matcher/Registry.h" -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" namespace mlir { diff --git a/mlir/include/mlir/Tools/mlir-reduce/MlirReduceMain.h b/mlir/include/mlir/Tools/mlir-reduce/MlirReduceMain.h index fccff9e..cb215b3 100644 --- a/mlir/include/mlir/Tools/mlir-reduce/MlirReduceMain.h +++ b/mlir/include/mlir/Tools/mlir-reduce/MlirReduceMain.h @@ -9,7 +9,7 @@ #ifndef MLIR_TOOLS_MLIR_REDUCE_MLIRREDUCEMAIN_H #define MLIR_TOOLS_MLIR_REDUCE_MLIRREDUCEMAIN_H -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" namespace mlir { diff --git a/mlir/include/mlir/Tools/mlir-translate/MlirTranslateMain.h b/mlir/include/mlir/Tools/mlir-translate/MlirTranslateMain.h index 4776987..c9bc58b 100644 --- a/mlir/include/mlir/Tools/mlir-translate/MlirTranslateMain.h +++ b/mlir/include/mlir/Tools/mlir-translate/MlirTranslateMain.h @@ -13,7 +13,7 @@ #ifndef MLIR_TOOLS_MLIRTRANSLATE_MLIRTRANSLATEMAIN_H #define MLIR_TOOLS_MLIRTRANSLATE_MLIRTRANSLATEMAIN_H -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "llvm/ADT/StringRef.h" namespace mlir { diff --git a/mlir/include/mlir/Tools/tblgen-lsp-server/TableGenLspServerMain.h b/mlir/include/mlir/Tools/tblgen-lsp-server/TableGenLspServerMain.h index 6859300..f0799c5 100644 --- a/mlir/include/mlir/Tools/tblgen-lsp-server/TableGenLspServerMain.h +++ b/mlir/include/mlir/Tools/tblgen-lsp-server/TableGenLspServerMain.h @@ -13,12 +13,13 @@ #ifndef MLIR_TOOLS_TBLGENLSPSERVER_TABLEGENLSPSERVERMAIN_H #define MLIR_TOOLS_TBLGENLSPSERVER_TABLEGENLSPSERVERMAIN_H -namespace mlir { +namespace llvm { struct LogicalResult; +} // namespace llvm +namespace mlir { /// Implementation for tools like `tblgen-lsp-server`. -LogicalResult TableGenLspServerMain(int argc, char **argv); - +llvm::LogicalResult TableGenLspServerMain(int argc, char **argv); } // namespace mlir #endif // MLIR_TOOLS_TBLGENLSPSERVER_TABLEGENLSPSERVERMAIN_H diff --git a/mlir/include/mlir/Transforms/HomomorphismSimplification.h b/mlir/include/mlir/Transforms/HomomorphismSimplification.h index d273260..386a315 100644 --- a/mlir/include/mlir/Transforms/HomomorphismSimplification.h +++ b/mlir/include/mlir/Transforms/HomomorphismSimplification.h @@ -13,7 +13,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/Value.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/SmallVector.h" #include "llvm/Support/Casting.h" #include <iterator> diff --git a/mlir/include/mlir/Transforms/Inliner.h b/mlir/include/mlir/Transforms/Inliner.h index 073b83f..ec77319 100644 --- a/mlir/include/mlir/Transforms/Inliner.h +++ b/mlir/include/mlir/Transforms/Inliner.h @@ -17,7 +17,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Pass/AnalysisManager.h" #include "mlir/Pass/PassManager.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/StringMap.h" namespace mlir { diff --git a/mlir/include/mlir/Transforms/LocationSnapshot.h b/mlir/include/mlir/Transforms/LocationSnapshot.h index a2b4200..ccfdbac 100644 --- a/mlir/include/mlir/Transforms/LocationSnapshot.h +++ b/mlir/include/mlir/Transforms/LocationSnapshot.h @@ -21,7 +21,6 @@ namespace mlir { class Location; -struct LogicalResult; class Operation; class OpPrintingFlags; class Pass; diff --git a/mlir/include/mlir/Transforms/SROA.h b/mlir/include/mlir/Transforms/SROA.h index d48f809..88e4bab 100644 --- a/mlir/include/mlir/Transforms/SROA.h +++ b/mlir/include/mlir/Transforms/SROA.h @@ -10,7 +10,7 @@ #define MLIR_TRANSFORMS_SROA_H #include "mlir/Interfaces/MemorySlotInterfaces.h" -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "llvm/ADT/Statistic.h" namespace mlir { diff --git a/mlir/lib/Analysis/AliasAnalysis/LocalAliasAnalysis.cpp b/mlir/lib/Analysis/AliasAnalysis/LocalAliasAnalysis.cpp index b22f3fd..6cece46 100644 --- a/mlir/lib/Analysis/AliasAnalysis/LocalAliasAnalysis.cpp +++ b/mlir/lib/Analysis/AliasAnalysis/LocalAliasAnalysis.cpp @@ -22,7 +22,6 @@ #include "mlir/Interfaces/SideEffectInterfaces.h" #include "mlir/Interfaces/ViewLikeInterface.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/Casting.h" #include <cassert> #include <optional> diff --git a/mlir/lib/Analysis/DataFlow/ConstantPropagationAnalysis.cpp b/mlir/lib/Analysis/DataFlow/ConstantPropagationAnalysis.cpp index 99c0419..16799d3 100644 --- a/mlir/lib/Analysis/DataFlow/ConstantPropagationAnalysis.cpp +++ b/mlir/lib/Analysis/DataFlow/ConstantPropagationAnalysis.cpp @@ -13,7 +13,6 @@ #include "mlir/IR/Operation.h" #include "mlir/IR/Value.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/Support/Casting.h" #include "llvm/Support/Debug.h" diff --git a/mlir/lib/Analysis/DataFlow/DeadCodeAnalysis.cpp b/mlir/lib/Analysis/DataFlow/DeadCodeAnalysis.cpp index 64df74e..fab2bd8 100644 --- a/mlir/lib/Analysis/DataFlow/DeadCodeAnalysis.cpp +++ b/mlir/lib/Analysis/DataFlow/DeadCodeAnalysis.cpp @@ -21,7 +21,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Interfaces/ControlFlowInterfaces.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/Casting.h" #include <cassert> #include <optional> diff --git a/mlir/lib/Analysis/DataFlow/DenseAnalysis.cpp b/mlir/lib/Analysis/DataFlow/DenseAnalysis.cpp index d4b9134..9894810 100644 --- a/mlir/lib/Analysis/DataFlow/DenseAnalysis.cpp +++ b/mlir/lib/Analysis/DataFlow/DenseAnalysis.cpp @@ -16,7 +16,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Interfaces/ControlFlowInterfaces.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/Support/Casting.h" #include <cassert> diff --git a/mlir/lib/Analysis/DataFlow/SparseAnalysis.cpp b/mlir/lib/Analysis/DataFlow/SparseAnalysis.cpp index b47bba1..ad956b7 100644 --- a/mlir/lib/Analysis/DataFlow/SparseAnalysis.cpp +++ b/mlir/lib/Analysis/DataFlow/SparseAnalysis.cpp @@ -18,7 +18,6 @@ #include "mlir/Interfaces/CallInterfaces.h" #include "mlir/Interfaces/ControlFlowInterfaces.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/Support/Casting.h" #include <cassert> diff --git a/mlir/lib/Analysis/DataFlowFramework.cpp b/mlir/lib/Analysis/DataFlowFramework.cpp index 5ef24f2..d0e827a 100644 --- a/mlir/lib/Analysis/DataFlowFramework.cpp +++ b/mlir/lib/Analysis/DataFlowFramework.cpp @@ -10,7 +10,6 @@ #include "mlir/IR/Location.h" #include "mlir/IR/Operation.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/iterator.h" #include "llvm/Config/abi-breaking.h" #include "llvm/Support/Casting.h" diff --git a/mlir/lib/Analysis/FlatLinearValueConstraints.cpp b/mlir/lib/Analysis/FlatLinearValueConstraints.cpp index 18047a6..746cff5 100644 --- a/mlir/lib/Analysis/FlatLinearValueConstraints.cpp +++ b/mlir/lib/Analysis/FlatLinearValueConstraints.cpp @@ -16,7 +16,6 @@ #include "mlir/IR/Builders.h" #include "mlir/IR/IntegerSet.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallPtrSet.h" #include "llvm/ADT/SmallVector.h" diff --git a/mlir/lib/AsmParser/AffineParser.cpp b/mlir/lib/AsmParser/AffineParser.cpp index 54c2a1f..1797611 100644 --- a/mlir/lib/AsmParser/AffineParser.cpp +++ b/mlir/lib/AsmParser/AffineParser.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/IntegerSet.h" #include "mlir/IR/OpImplementation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/ErrorHandling.h" #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/SourceMgr.h" diff --git a/mlir/lib/AsmParser/AsmParserState.cpp b/mlir/lib/AsmParser/AsmParserState.cpp index 47cfb52..589bf6a 100644 --- a/mlir/lib/AsmParser/AsmParserState.cpp +++ b/mlir/lib/AsmParser/AsmParserState.cpp @@ -13,7 +13,6 @@ #include "mlir/IR/Types.h" #include "mlir/IR/Value.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/StringExtras.h" diff --git a/mlir/lib/AsmParser/DialectSymbolParser.cpp b/mlir/lib/AsmParser/DialectSymbolParser.cpp index 80cce7e..9f4a87a 100644 --- a/mlir/lib/AsmParser/DialectSymbolParser.cpp +++ b/mlir/lib/AsmParser/DialectSymbolParser.cpp @@ -23,7 +23,6 @@ #include "mlir/IR/DialectImplementation.h" #include "mlir/IR/MLIRContext.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/SourceMgr.h" #include <cassert> diff --git a/mlir/lib/AsmParser/LocationParser.cpp b/mlir/lib/AsmParser/LocationParser.cpp index fed65c0..1365da0 100644 --- a/mlir/lib/AsmParser/LocationParser.cpp +++ b/mlir/lib/AsmParser/LocationParser.cpp @@ -12,7 +12,6 @@ #include "mlir/IR/BuiltinAttributes.h" #include "mlir/IR/Location.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" using namespace mlir; using namespace mlir::detail; diff --git a/mlir/lib/AsmParser/Parser.cpp b/mlir/lib/AsmParser/Parser.cpp index 1b8b4ba..7181f13 100644 --- a/mlir/lib/AsmParser/Parser.cpp +++ b/mlir/lib/AsmParser/Parser.cpp @@ -34,7 +34,6 @@ #include "mlir/IR/Verifier.h" #include "mlir/IR/Visitors.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/TypeID.h" #include "llvm/ADT/APFloat.h" #include "llvm/ADT/DenseMap.h" diff --git a/mlir/lib/AsmParser/TypeParser.cpp b/mlir/lib/AsmParser/TypeParser.cpp index 5da931b..0b46c96 100644 --- a/mlir/lib/AsmParser/TypeParser.cpp +++ b/mlir/lib/AsmParser/TypeParser.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/TensorEncoding.h" #include "mlir/IR/Types.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include <cassert> #include <cstdint> diff --git a/mlir/lib/Bytecode/Reader/BytecodeReader.cpp b/mlir/lib/Bytecode/Reader/BytecodeReader.cpp index dd1e4ab..f767740 100644 --- a/mlir/lib/Bytecode/Reader/BytecodeReader.cpp +++ b/mlir/lib/Bytecode/Reader/BytecodeReader.cpp @@ -17,7 +17,6 @@ #include "mlir/IR/Verifier.h" #include "mlir/IR/Visitors.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/ScopeExit.h" #include "llvm/ADT/StringExtras.h" diff --git a/mlir/lib/Bytecode/Writer/BytecodeWriter.cpp b/mlir/lib/Bytecode/Writer/BytecodeWriter.cpp index 9493a6c..449d754 100644 --- a/mlir/lib/Bytecode/Writer/BytecodeWriter.cpp +++ b/mlir/lib/Bytecode/Writer/BytecodeWriter.cpp @@ -14,7 +14,6 @@ #include "mlir/IR/Attributes.h" #include "mlir/IR/Diagnostics.h" #include "mlir/IR/OpImplementation.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/CachedHashString.h" #include "llvm/ADT/MapVector.h" diff --git a/mlir/lib/CAPI/IR/BuiltinTypes.cpp b/mlir/lib/CAPI/IR/BuiltinTypes.cpp index c94c070..01bb71d 100644 --- a/mlir/lib/CAPI/IR/BuiltinTypes.cpp +++ b/mlir/lib/CAPI/IR/BuiltinTypes.cpp @@ -16,7 +16,6 @@ #include "mlir/IR/AffineMap.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/Types.h" -#include "mlir/Support/LogicalResult.h" #include <algorithm> diff --git a/mlir/lib/Conversion/ArithToAMDGPU/ArithToAMDGPU.cpp b/mlir/lib/Conversion/ArithToAMDGPU/ArithToAMDGPU.cpp index 3d3ff00..58764ad 100644 --- a/mlir/lib/Conversion/ArithToAMDGPU/ArithToAMDGPU.cpp +++ b/mlir/lib/Conversion/ArithToAMDGPU/ArithToAMDGPU.cpp @@ -16,7 +16,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" namespace mlir { diff --git a/mlir/lib/Conversion/ArithToEmitC/ArithToEmitC.cpp b/mlir/lib/Conversion/ArithToEmitC/ArithToEmitC.cpp index 93717e3..7c6fdab 100644 --- a/mlir/lib/Conversion/ArithToEmitC/ArithToEmitC.cpp +++ b/mlir/lib/Conversion/ArithToEmitC/ArithToEmitC.cpp @@ -17,7 +17,6 @@ #include "mlir/Dialect/EmitC/IR/EmitC.h" #include "mlir/IR/BuiltinAttributes.h" #include "mlir/IR/BuiltinTypes.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" using namespace mlir; diff --git a/mlir/lib/Conversion/BufferizationToMemRef/BufferizationToMemRef.cpp b/mlir/lib/Conversion/BufferizationToMemRef/BufferizationToMemRef.cpp index 810f82f..2aae39f 100644 --- a/mlir/lib/Conversion/BufferizationToMemRef/BufferizationToMemRef.cpp +++ b/mlir/lib/Conversion/BufferizationToMemRef/BufferizationToMemRef.cpp @@ -21,7 +21,6 @@ #include "mlir/Dialect/SCF/IR/SCF.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" namespace mlir { diff --git a/mlir/lib/Conversion/ControlFlowToSPIRV/ControlFlowToSPIRV.cpp b/mlir/lib/Conversion/ControlFlowToSPIRV/ControlFlowToSPIRV.cpp index 995eecf..f96bfd6 100644 --- a/mlir/lib/Conversion/ControlFlowToSPIRV/ControlFlowToSPIRV.cpp +++ b/mlir/lib/Conversion/ControlFlowToSPIRV/ControlFlowToSPIRV.cpp @@ -19,7 +19,6 @@ #include "mlir/Dialect/SPIRV/Utils/LayoutUtils.h" #include "mlir/IR/AffineMap.h" #include "mlir/IR/PatternMatch.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/Support/Debug.h" #include "llvm/Support/FormatVariadic.h" diff --git a/mlir/lib/Conversion/FuncToLLVM/FuncToLLVM.cpp b/mlir/lib/Conversion/FuncToLLVM/FuncToLLVM.cpp index efb8046..059acb2 100644 --- a/mlir/lib/Conversion/FuncToLLVM/FuncToLLVM.cpp +++ b/mlir/lib/Conversion/FuncToLLVM/FuncToLLVM.cpp @@ -35,7 +35,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/SymbolTable.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "mlir/Transforms/Passes.h" #include "llvm/ADT/SmallVector.h" diff --git a/mlir/lib/Conversion/FuncToSPIRV/FuncToSPIRV.cpp b/mlir/lib/Conversion/FuncToSPIRV/FuncToSPIRV.cpp index a575551..4740b7c 100644 --- a/mlir/lib/Conversion/FuncToSPIRV/FuncToSPIRV.cpp +++ b/mlir/lib/Conversion/FuncToSPIRV/FuncToSPIRV.cpp @@ -18,7 +18,6 @@ #include "mlir/Dialect/SPIRV/Transforms/SPIRVConversion.h" #include "mlir/Dialect/SPIRV/Utils/LayoutUtils.h" #include "mlir/IR/AffineMap.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/Debug.h" #define DEBUG_TYPE "func-to-spirv-pattern" diff --git a/mlir/lib/Conversion/GPUToSPIRV/GPUToSPIRV.cpp b/mlir/lib/Conversion/GPUToSPIRV/GPUToSPIRV.cpp index d7885e0..98340bf 100644 --- a/mlir/lib/Conversion/GPUToSPIRV/GPUToSPIRV.cpp +++ b/mlir/lib/Conversion/GPUToSPIRV/GPUToSPIRV.cpp @@ -21,7 +21,6 @@ #include "mlir/Dialect/SPIRV/Transforms/SPIRVConversion.h" #include "mlir/IR/BuiltinOps.h" #include "mlir/IR/Matchers.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include <optional> diff --git a/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp b/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp index 81b9f55..90b0d72 100644 --- a/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp +++ b/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp @@ -21,7 +21,6 @@ #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/MLIRContext.h" #include "mlir/IR/Visitors.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/Debug.h" #include <cassert> #include <optional> diff --git a/mlir/lib/Conversion/NVVMToLLVM/NVVMToLLVM.cpp b/mlir/lib/Conversion/NVVMToLLVM/NVVMToLLVM.cpp index d1d68e3..662ee9e 100644 --- a/mlir/lib/Conversion/NVVMToLLVM/NVVMToLLVM.cpp +++ b/mlir/lib/Conversion/NVVMToLLVM/NVVMToLLVM.cpp @@ -25,7 +25,6 @@ #include "mlir/IR/Value.h" #include "mlir/Pass/Pass.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/raw_ostream.h" #define DEBUG_TYPE "nvvm-to-llvm" diff --git a/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp b/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp index 885bb5a..da09384 100644 --- a/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp +++ b/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp @@ -20,7 +20,6 @@ #include "mlir/Dialect/SPIRV/Utils/LayoutUtils.h" #include "mlir/IR/BuiltinOps.h" #include "mlir/IR/PatternMatch.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/Support/Debug.h" #include "llvm/Support/FormatVariadic.h" diff --git a/mlir/lib/Conversion/TensorToLinalg/TensorToLinalg.cpp b/mlir/lib/Conversion/TensorToLinalg/TensorToLinalg.cpp index 0458073..5bb79d4 100644 --- a/mlir/lib/Conversion/TensorToLinalg/TensorToLinalg.cpp +++ b/mlir/lib/Conversion/TensorToLinalg/TensorToLinalg.cpp @@ -14,7 +14,6 @@ #include "mlir/Dialect/Linalg/Transforms/Transforms.h" #include "mlir/Dialect/Tensor/IR/Tensor.h" #include "mlir/IR/AffineMap.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/Debug.h" #define DEBUG_TYPE "tensor-to-linalg-pattern" diff --git a/mlir/lib/Conversion/TensorToSPIRV/TensorToSPIRV.cpp b/mlir/lib/Conversion/TensorToSPIRV/TensorToSPIRV.cpp index 373952c..0fb5862 100644 --- a/mlir/lib/Conversion/TensorToSPIRV/TensorToSPIRV.cpp +++ b/mlir/lib/Conversion/TensorToSPIRV/TensorToSPIRV.cpp @@ -18,7 +18,6 @@ #include "mlir/Dialect/SPIRV/Utils/LayoutUtils.h" #include "mlir/Dialect/Tensor/IR/Tensor.h" #include "mlir/IR/AffineMap.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/Debug.h" #define DEBUG_TYPE "tensor-to-spirv-pattern" diff --git a/mlir/lib/Conversion/VectorToGPU/VectorToGPU.cpp b/mlir/lib/Conversion/VectorToGPU/VectorToGPU.cpp index 0fd91b27..0150ff6 100644 --- a/mlir/lib/Conversion/VectorToGPU/VectorToGPU.cpp +++ b/mlir/lib/Conversion/VectorToGPU/VectorToGPU.cpp @@ -31,7 +31,6 @@ #include "mlir/IR/BuiltinOps.h" #include "mlir/IR/Region.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include "mlir/Transforms/Passes.h" #include "llvm/ADT/STLExtras.h" diff --git a/mlir/lib/Conversion/VectorToSPIRV/VectorToSPIRV.cpp b/mlir/lib/Conversion/VectorToSPIRV/VectorToSPIRV.cpp index 8baa31a..c936329 100644 --- a/mlir/lib/Conversion/VectorToSPIRV/VectorToSPIRV.cpp +++ b/mlir/lib/Conversion/VectorToSPIRV/VectorToSPIRV.cpp @@ -25,7 +25,6 @@ #include "mlir/IR/Matchers.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" diff --git a/mlir/lib/Dialect/Affine/Transforms/AffineScalarReplacement.cpp b/mlir/lib/Dialect/Affine/Transforms/AffineScalarReplacement.cpp index 707bba2..16c0d3d 100644 --- a/mlir/lib/Dialect/Affine/Transforms/AffineScalarReplacement.cpp +++ b/mlir/lib/Dialect/Affine/Transforms/AffineScalarReplacement.cpp @@ -17,7 +17,6 @@ #include "mlir/Dialect/Affine/Utils.h" #include "mlir/Dialect/Func/IR/FuncOps.h" #include "mlir/IR/Dominance.h" -#include "mlir/Support/LogicalResult.h" #include <algorithm> namespace mlir { diff --git a/mlir/lib/Dialect/Arith/IR/ArithOps.cpp b/mlir/lib/Dialect/Arith/IR/ArithOps.cpp index 0e7beff..aa5eb95 100644 --- a/mlir/lib/Dialect/Arith/IR/ArithOps.cpp +++ b/mlir/lib/Dialect/Arith/IR/ArithOps.cpp @@ -21,7 +21,6 @@ #include "mlir/IR/OpImplementation.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APFloat.h" #include "llvm/ADT/APInt.h" diff --git a/mlir/lib/Dialect/Arith/Transforms/EmulateNarrowType.cpp b/mlir/lib/Dialect/Arith/Transforms/EmulateNarrowType.cpp index e0e1385..1c1f65e 100644 --- a/mlir/lib/Dialect/Arith/Transforms/EmulateNarrowType.cpp +++ b/mlir/lib/Dialect/Arith/Transforms/EmulateNarrowType.cpp @@ -15,7 +15,6 @@ #include "mlir/Dialect/Func/Transforms/FuncConversions.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/ADT/APInt.h" #include "llvm/Support/FormatVariadic.h" diff --git a/mlir/lib/Dialect/Arith/Transforms/EmulateWideInt.cpp b/mlir/lib/Dialect/Arith/Transforms/EmulateWideInt.cpp index 8a4080e..9c91dd7 100644 --- a/mlir/lib/Dialect/Arith/Transforms/EmulateWideInt.cpp +++ b/mlir/lib/Dialect/Arith/Transforms/EmulateWideInt.cpp @@ -16,7 +16,6 @@ #include "mlir/Dialect/Vector/IR/VectorOps.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/ADT/APInt.h" #include "llvm/Support/FormatVariadic.h" diff --git a/mlir/lib/Dialect/Arith/Transforms/IntNarrowing.cpp b/mlir/lib/Dialect/Arith/Transforms/IntNarrowing.cpp index f87f3d6..e2d42e9 100644 --- a/mlir/lib/Dialect/Arith/Transforms/IntNarrowing.cpp +++ b/mlir/lib/Dialect/Arith/Transforms/IntNarrowing.cpp @@ -21,7 +21,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/ValueBoundsOpInterface.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallVector.h" diff --git a/mlir/lib/Dialect/ArmNeon/Transforms/LowerContractionToSMMLAPattern.cpp b/mlir/lib/Dialect/ArmNeon/Transforms/LowerContractionToSMMLAPattern.cpp index 3635cd3..2c0c84d 100644 --- a/mlir/lib/Dialect/ArmNeon/Transforms/LowerContractionToSMMLAPattern.cpp +++ b/mlir/lib/Dialect/ArmNeon/Transforms/LowerContractionToSMMLAPattern.cpp @@ -20,7 +20,6 @@ #include "mlir/Dialect/Vector/IR/VectorOps.h" #include "mlir/IR/AffineMap.h" #include "mlir/IR/PatternMatch.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #define DEBUG_TYPE "lower-contract-to-arm-neon" diff --git a/mlir/lib/Dialect/Bufferization/Transforms/LowerDeallocations.cpp b/mlir/lib/Dialect/Bufferization/Transforms/LowerDeallocations.cpp index 982d955..7fb4691 100644 --- a/mlir/lib/Dialect/Bufferization/Transforms/LowerDeallocations.cpp +++ b/mlir/lib/Dialect/Bufferization/Transforms/LowerDeallocations.cpp @@ -19,7 +19,6 @@ #include "mlir/Dialect/SCF/IR/SCF.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" namespace mlir { diff --git a/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp b/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp index 3abaa3b..7bc2668 100644 --- a/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp +++ b/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp @@ -29,7 +29,6 @@ #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Interfaces/SideEffectInterfaces.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/InliningUtils.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/TypeSwitch.h" diff --git a/mlir/lib/Dialect/GPU/Pipelines/GPUToNVVMPipeline.cpp b/mlir/lib/Dialect/GPU/Pipelines/GPUToNVVMPipeline.cpp index f457303..fb44075 100644 --- a/mlir/lib/Dialect/GPU/Pipelines/GPUToNVVMPipeline.cpp +++ b/mlir/lib/Dialect/GPU/Pipelines/GPUToNVVMPipeline.cpp @@ -38,7 +38,6 @@ using namespace mlir; -#if LLVM_HAS_NVPTX_TARGET namespace { //===----------------------------------------------------------------------===// @@ -126,5 +125,3 @@ void mlir::gpu::registerGPUToNVVMPipeline() { "code.", buildLowerToNVVMPassPipeline); } - -#endif // LLVM_HAS_NVPTX_TARGET diff --git a/mlir/lib/Dialect/GPU/Transforms/ModuleToBinary.cpp b/mlir/lib/Dialect/GPU/Transforms/ModuleToBinary.cpp index 1e7596e..adae3be 100644 --- a/mlir/lib/Dialect/GPU/Transforms/ModuleToBinary.cpp +++ b/mlir/lib/Dialect/GPU/Transforms/ModuleToBinary.cpp @@ -38,25 +38,10 @@ class GpuModuleToBinaryPass : public impl::GpuModuleToBinaryPassBase<GpuModuleToBinaryPass> { public: using Base::Base; - void getDependentDialects(DialectRegistry ®istry) const override; void runOnOperation() final; }; } // namespace -void GpuModuleToBinaryPass::getDependentDialects( - DialectRegistry ®istry) const { - // Register all GPU related translations. - registry.insert<gpu::GPUDialect>(); - registry.insert<LLVM::LLVMDialect>(); -#if LLVM_HAS_NVPTX_TARGET - registry.insert<NVVM::NVVMDialect>(); -#endif -#if MLIR_ENABLE_ROCM_CONVERSIONS - registry.insert<ROCDL::ROCDLDialect>(); -#endif - registry.insert<spirv::SPIRVDialect>(); -} - void GpuModuleToBinaryPass::runOnOperation() { RewritePatternSet patterns(&getContext()); auto targetFormat = diff --git a/mlir/lib/Dialect/GPU/Transforms/SubgroupReduceLowering.cpp b/mlir/lib/Dialect/GPU/Transforms/SubgroupReduceLowering.cpp index b00c65c..561a7e5 100644 --- a/mlir/lib/Dialect/GPU/Transforms/SubgroupReduceLowering.cpp +++ b/mlir/lib/Dialect/GPU/Transforms/SubgroupReduceLowering.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/Location.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/FormatVariadic.h" #include "llvm/Support/MathExtras.h" #include <cassert> diff --git a/mlir/lib/Dialect/IRDL/IR/IRDL.cpp b/mlir/lib/Dialect/IRDL/IR/IRDL.cpp index 1f5584f..c5c44d9 100644 --- a/mlir/lib/Dialect/IRDL/IR/IRDL.cpp +++ b/mlir/lib/Dialect/IRDL/IR/IRDL.cpp @@ -17,7 +17,6 @@ #include "mlir/IR/OpImplementation.h" #include "mlir/IR/Operation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/TypeSwitch.h" #include "llvm/IR/Metadata.h" diff --git a/mlir/lib/Dialect/IRDL/IRDLLoading.cpp b/mlir/lib/Dialect/IRDL/IRDLLoading.cpp index 5f623e8..9a2754f 100644 --- a/mlir/lib/Dialect/IRDL/IRDLLoading.cpp +++ b/mlir/lib/Dialect/IRDL/IRDLLoading.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/BuiltinOps.h" #include "mlir/IR/ExtensibleDialect.h" #include "mlir/IR/OperationSupport.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallPtrSet.h" #include "llvm/Support/SMLoc.h" diff --git a/mlir/lib/Dialect/IRDL/IRDLVerifiers.cpp b/mlir/lib/Dialect/IRDL/IRDLVerifiers.cpp index 05dc154..50f04d0 100644 --- a/mlir/lib/Dialect/IRDL/IRDLVerifiers.cpp +++ b/mlir/lib/Dialect/IRDL/IRDLVerifiers.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/Location.h" #include "mlir/IR/Region.h" #include "mlir/IR/Value.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/FormatVariadic.h" using namespace mlir; diff --git a/mlir/lib/Dialect/LLVMIR/IR/BasicPtxBuilderInterface.cpp b/mlir/lib/Dialect/LLVMIR/IR/BasicPtxBuilderInterface.cpp index 3dacc70..961944c 100644 --- a/mlir/lib/Dialect/LLVMIR/IR/BasicPtxBuilderInterface.cpp +++ b/mlir/lib/Dialect/LLVMIR/IR/BasicPtxBuilderInterface.cpp @@ -12,7 +12,6 @@ //===----------------------------------------------------------------------===// #include "mlir/Dialect/LLVMIR/BasicPtxBuilderInterface.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "ptx-builder" #define DBGS() (llvm::dbgs() << "[" DEBUG_TYPE "]: ") diff --git a/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp b/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp index 94197e4..036a9a1 100644 --- a/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp +++ b/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp @@ -28,7 +28,6 @@ #include "mlir/IR/Operation.h" #include "mlir/IR/OperationSupport.h" #include "mlir/IR/Types.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/TypeSwitch.h" #include "llvm/AsmParser/Parser.h" diff --git a/mlir/lib/Dialect/Linalg/Transforms/MeshShardingInterfaceImpl.cpp b/mlir/lib/Dialect/Linalg/Transforms/MeshShardingInterfaceImpl.cpp index 24001c5..36b6088 100644 --- a/mlir/lib/Dialect/Linalg/Transforms/MeshShardingInterfaceImpl.cpp +++ b/mlir/lib/Dialect/Linalg/Transforms/MeshShardingInterfaceImpl.cpp @@ -30,7 +30,6 @@ #include "mlir/IR/SymbolTable.h" #include "mlir/IR/Value.h" #include "mlir/Interfaces/TilingInterface.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallVector.h" diff --git a/mlir/lib/Dialect/Linalg/Transforms/TransposeConv2D.cpp b/mlir/lib/Dialect/Linalg/Transforms/TransposeConv2D.cpp index 9e0829e..1294043 100644 --- a/mlir/lib/Dialect/Linalg/Transforms/TransposeConv2D.cpp +++ b/mlir/lib/Dialect/Linalg/Transforms/TransposeConv2D.cpp @@ -13,7 +13,6 @@ #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/ValueRange.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include "llvm/ADT/SmallVector.h" diff --git a/mlir/lib/Dialect/MemRef/IR/MemRefMemorySlot.cpp b/mlir/lib/Dialect/MemRef/IR/MemRefMemorySlot.cpp index 847c2f3..f630c48 100644 --- a/mlir/lib/Dialect/MemRef/IR/MemRefMemorySlot.cpp +++ b/mlir/lib/Dialect/MemRef/IR/MemRefMemorySlot.cpp @@ -20,7 +20,6 @@ #include "mlir/IR/Value.h" #include "mlir/Interfaces/InferTypeOpInterface.h" #include "mlir/Interfaces/MemorySlotInterfaces.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/TypeSwitch.h" #include "llvm/Support/ErrorHandling.h" diff --git a/mlir/lib/Dialect/MemRef/Transforms/EmulateNarrowType.cpp b/mlir/lib/Dialect/MemRef/Transforms/EmulateNarrowType.cpp index f140c04..20ce1b1 100644 --- a/mlir/lib/Dialect/MemRef/Transforms/EmulateNarrowType.cpp +++ b/mlir/lib/Dialect/MemRef/Transforms/EmulateNarrowType.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/Builders.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/OpDefinition.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/Support/FormatVariadic.h" #include "llvm/Support/MathExtras.h" diff --git a/mlir/lib/Dialect/Mesh/IR/MeshOps.cpp b/mlir/lib/Dialect/Mesh/IR/MeshOps.cpp index ec1acbb..75fceee 100644 --- a/mlir/lib/Dialect/Mesh/IR/MeshOps.cpp +++ b/mlir/lib/Dialect/Mesh/IR/MeshOps.cpp @@ -23,7 +23,6 @@ #include "mlir/IR/Value.h" #include "mlir/Interfaces/ViewLikeInterface.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallSet.h" diff --git a/mlir/lib/Dialect/Mesh/Interfaces/ShardingInterface.cpp b/mlir/lib/Dialect/Mesh/Interfaces/ShardingInterface.cpp index a233847..bcd0e15 100644 --- a/mlir/lib/Dialect/Mesh/Interfaces/ShardingInterface.cpp +++ b/mlir/lib/Dialect/Mesh/Interfaces/ShardingInterface.cpp @@ -13,7 +13,6 @@ #include "mlir/IR/AffineMap.h" #include "mlir/IR/IRMapping.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallSet.h" diff --git a/mlir/lib/Dialect/Mesh/Transforms/ShardingPropagation.cpp b/mlir/lib/Dialect/Mesh/Transforms/ShardingPropagation.cpp index 870ac4a..511c910 100644 --- a/mlir/lib/Dialect/Mesh/Transforms/ShardingPropagation.cpp +++ b/mlir/lib/Dialect/Mesh/Transforms/ShardingPropagation.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/Verifier.h" #include "mlir/Interfaces/FunctionInterfaces.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallVector.h" #include "llvm/ADT/iterator_range.h" diff --git a/mlir/lib/Dialect/Mesh/Transforms/Simplifications.cpp b/mlir/lib/Dialect/Mesh/Transforms/Simplifications.cpp index a29551a..4c14b1c 100644 --- a/mlir/lib/Dialect/Mesh/Transforms/Simplifications.cpp +++ b/mlir/lib/Dialect/Mesh/Transforms/Simplifications.cpp @@ -14,7 +14,6 @@ #include "mlir/IR/ImplicitLocOpBuilder.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/SymbolTable.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallVector.h" #include <numeric> diff --git a/mlir/lib/Dialect/Mesh/Transforms/Spmdization.cpp b/mlir/lib/Dialect/Mesh/Transforms/Spmdization.cpp index 4ecc897..1df3cf6 100644 --- a/mlir/lib/Dialect/Mesh/Transforms/Spmdization.cpp +++ b/mlir/lib/Dialect/Mesh/Transforms/Spmdization.cpp @@ -28,7 +28,6 @@ #include "mlir/Interfaces/FunctionInterfaces.h" #include "mlir/Pass/Pass.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APInt.h" #include "llvm/ADT/DenseSet.h" #include "llvm/ADT/STLExtras.h" diff --git a/mlir/lib/Dialect/NVGPU/Transforms/MmaSyncTF32Transform.cpp b/mlir/lib/Dialect/NVGPU/Transforms/MmaSyncTF32Transform.cpp index 8163f42..6222cc7 100644 --- a/mlir/lib/Dialect/NVGPU/Transforms/MmaSyncTF32Transform.cpp +++ b/mlir/lib/Dialect/NVGPU/Transforms/MmaSyncTF32Transform.cpp @@ -19,7 +19,6 @@ #include "mlir/Dialect/NVGPU/IR/NVGPUDialect.h" #include "mlir/Dialect/Vector/IR/VectorOps.h" #include "mlir/Interfaces/SideEffectInterfaces.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/Support/MathExtras.h" diff --git a/mlir/lib/Dialect/NVGPU/Transforms/OptimizeSharedMemory.cpp b/mlir/lib/Dialect/NVGPU/Transforms/OptimizeSharedMemory.cpp index 693bb53..31ffacb 100644 --- a/mlir/lib/Dialect/NVGPU/Transforms/OptimizeSharedMemory.cpp +++ b/mlir/lib/Dialect/NVGPU/Transforms/OptimizeSharedMemory.cpp @@ -20,7 +20,6 @@ #include "mlir/Dialect/NVGPU/Transforms/Utils.h" #include "mlir/Dialect/Vector/IR/VectorOps.h" #include "mlir/Interfaces/SideEffectInterfaces.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/Support/MathExtras.h" @@ -146,7 +145,7 @@ getShmReadAndWriteOps(Operation *parentOp, Value shmMemRef, return success(); } -mlir::LogicalResult +llvm::LogicalResult mlir::nvgpu::optimizeSharedMemoryReadsAndWrites(Operation *parentOp, Value memrefValue) { auto memRefType = dyn_cast<MemRefType>(memrefValue.getType()); diff --git a/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp b/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp index b13c8d1..abbd857 100644 --- a/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp +++ b/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp @@ -38,7 +38,6 @@ #include "mlir/Dialect/OpenMP/OpenMPOpsEnums.cpp.inc" #include "mlir/Dialect/OpenMP/OpenMPOpsInterfaces.cpp.inc" #include "mlir/Dialect/OpenMP/OpenMPTypeInterfaces.cpp.inc" -#include "mlir/Support/LogicalResult.h" using namespace mlir; using namespace mlir::omp; diff --git a/mlir/lib/Dialect/Polynomial/IR/Polynomial.cpp b/mlir/lib/Dialect/Polynomial/IR/Polynomial.cpp index e85bced..650a369 100644 --- a/mlir/lib/Dialect/Polynomial/IR/Polynomial.cpp +++ b/mlir/lib/Dialect/Polynomial/IR/Polynomial.cpp @@ -8,7 +8,6 @@ #include "mlir/Dialect/Polynomial/IR/Polynomial.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/SmallVector.h" namespace mlir { diff --git a/mlir/lib/Dialect/Polynomial/IR/PolynomialAttributes.cpp b/mlir/lib/Dialect/Polynomial/IR/PolynomialAttributes.cpp index cc7d317..93c7f9e 100644 --- a/mlir/lib/Dialect/Polynomial/IR/PolynomialAttributes.cpp +++ b/mlir/lib/Dialect/Polynomial/IR/PolynomialAttributes.cpp @@ -9,7 +9,6 @@ #include "mlir/Dialect/Polynomial/IR/Polynomial.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/SmallVector.h" #include "llvm/ADT/StringExtras.h" #include "llvm/ADT/StringRef.h" diff --git a/mlir/lib/Dialect/Polynomial/IR/PolynomialDialect.cpp b/mlir/lib/Dialect/Polynomial/IR/PolynomialDialect.cpp index 825b80d..7f8ba06 100644 --- a/mlir/lib/Dialect/Polynomial/IR/PolynomialDialect.cpp +++ b/mlir/lib/Dialect/Polynomial/IR/PolynomialDialect.cpp @@ -18,7 +18,6 @@ #include "mlir/IR/Dialect.h" #include "mlir/IR/PatternMatch.h" #include "mlir/Interfaces/InferTypeOpInterface.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APInt.h" #include "llvm/ADT/TypeSwitch.h" diff --git a/mlir/lib/Dialect/Polynomial/IR/PolynomialOps.cpp b/mlir/lib/Dialect/Polynomial/IR/PolynomialOps.cpp index 40fa97d..2ba13bb 100644 --- a/mlir/lib/Dialect/Polynomial/IR/PolynomialOps.cpp +++ b/mlir/lib/Dialect/Polynomial/IR/PolynomialOps.cpp @@ -15,7 +15,6 @@ #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/Dialect.h" #include "mlir/IR/PatternMatch.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APInt.h" using namespace mlir; diff --git a/mlir/lib/Dialect/Quant/IR/QuantDialectBytecode.cpp b/mlir/lib/Dialect/Quant/IR/QuantDialectBytecode.cpp index ee6be25..c0c00fb 100644 --- a/mlir/lib/Dialect/Quant/IR/QuantDialectBytecode.cpp +++ b/mlir/lib/Dialect/Quant/IR/QuantDialectBytecode.cpp @@ -12,7 +12,6 @@ #include "mlir/Dialect/Quant/QuantOps.h" #include "mlir/Dialect/Quant/QuantTypes.h" #include "mlir/IR/Diagnostics.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APFloat.h" #include "llvm/ADT/SmallVector.h" #include "llvm/ADT/TypeSwitch.h" diff --git a/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp b/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp index 3841562..d9bc05a 100644 --- a/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp +++ b/mlir/lib/Dialect/SPIRV/IR/SPIRVOps.cpp @@ -29,7 +29,6 @@ #include "mlir/IR/Operation.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/FunctionImplementation.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APFloat.h" #include "llvm/ADT/APInt.h" #include "llvm/ADT/ArrayRef.h" diff --git a/mlir/lib/Dialect/SPIRV/Transforms/SPIRVWebGPUTransforms.cpp b/mlir/lib/Dialect/SPIRV/Transforms/SPIRVWebGPUTransforms.cpp index 5d4dd5b..d75c855 100644 --- a/mlir/lib/Dialect/SPIRV/Transforms/SPIRVWebGPUTransforms.cpp +++ b/mlir/lib/Dialect/SPIRV/Transforms/SPIRVWebGPUTransforms.cpp @@ -17,7 +17,6 @@ #include "mlir/IR/Location.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" diff --git a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp index 288cbac..4337787 100644 --- a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp +++ b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp @@ -955,7 +955,7 @@ LogicalResult tosa::ReshapeOp::inferReturnTypeComponents( return success(); } -mlir::LogicalResult tosa::ReshapeOp::verify() { +llvm::LogicalResult tosa::ReshapeOp::verify() { TensorType inputType = getInput1().getType(); RankedTensorType outputType = getType(); diff --git a/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp b/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp index 6eef2c5..403ac48 100644 --- a/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp +++ b/mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp @@ -20,7 +20,6 @@ #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/Matchers.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APFloat.h" #include "llvm/ADT/FloatingPointMode.h" #include "llvm/ADT/SmallVector.h" diff --git a/mlir/lib/Dialect/Transform/Interfaces/TransformInterfaces.cpp b/mlir/lib/Dialect/Transform/Interfaces/TransformInterfaces.cpp index ff02cb2..e2b314f9 100644 --- a/mlir/lib/Dialect/Transform/Interfaces/TransformInterfaces.cpp +++ b/mlir/lib/Dialect/Transform/Interfaces/TransformInterfaces.cpp @@ -12,7 +12,6 @@ #include "mlir/IR/Operation.h" #include "mlir/IR/PatternMatch.h" #include "mlir/Interfaces/CastInterfaces.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/ScopeExit.h" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorBitCast.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorBitCast.cpp index e5f11d8..d8c4939 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorBitCast.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorBitCast.cpp @@ -16,7 +16,6 @@ #include "mlir/Dialect/Vector/Utils/VectorUtils.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/PatternMatch.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "vector-bitcast-lowering" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorBroadcast.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorBroadcast.cpp index 44e3f76..32e7eb2 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorBroadcast.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorBroadcast.cpp @@ -32,7 +32,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/VectorInterfaces.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "vector-broadcast-lowering" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp index ba1c968..3a799ce 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp @@ -31,7 +31,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/VectorInterfaces.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "vector-contract-lowering" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorGather.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorGather.cpp index a0df03c..f1a5aa7 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorGather.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorGather.cpp @@ -31,7 +31,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/VectorInterfaces.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "vector-broadcast-lowering" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorInterleave.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorInterleave.cpp index f7e01c7..cab0f21 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorInterleave.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorInterleave.cpp @@ -16,7 +16,6 @@ #include "mlir/Dialect/Vector/Utils/VectorUtils.h" #include "mlir/IR/BuiltinTypes.h" #include "mlir/IR/PatternMatch.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "vector-interleave-lowering" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorScan.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorScan.cpp index c4e8ddc..92fddc1 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorScan.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorScan.cpp @@ -31,7 +31,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/VectorInterfaces.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "vector-broadcast-lowering" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp index 85c4c03..95ebd4e 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorShapeCast.cpp @@ -32,7 +32,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/VectorInterfaces.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "vector-shape-cast-lowering" diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorTranspose.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorTranspose.cpp index ca8a6f6..3c92b22 100644 --- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorTranspose.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorTranspose.cpp @@ -31,7 +31,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/VectorInterfaces.h" -#include "mlir/Support/LogicalResult.h" #define DEBUG_TYPE "lower-vector-transpose" diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorLinearize.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorLinearize.cpp index a1bb81e..4a3ae1b 100644 --- a/mlir/lib/Dialect/Vector/Transforms/VectorLinearize.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/VectorLinearize.cpp @@ -18,7 +18,6 @@ #include "mlir/IR/Operation.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/DialectConversion.h" #include "llvm/ADT/ArrayRef.h" #include <cstdint> diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorTransforms.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorTransforms.cpp index da3d964..c7d3022 100644 --- a/mlir/lib/Dialect/Vector/Transforms/VectorTransforms.cpp +++ b/mlir/lib/Dialect/Vector/Transforms/VectorTransforms.cpp @@ -38,7 +38,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" #include "mlir/Interfaces/VectorInterfaces.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/DenseSet.h" #include "llvm/ADT/MapVector.h" diff --git a/mlir/lib/IR/AffineMap.cpp b/mlir/lib/IR/AffineMap.cpp index e5993eb..62f5952 100644 --- a/mlir/lib/IR/AffineMap.cpp +++ b/mlir/lib/IR/AffineMap.cpp @@ -12,7 +12,6 @@ #include "mlir/IR/Builders.h" #include "mlir/IR/BuiltinAttributes.h" #include "mlir/IR/BuiltinTypes.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallBitVector.h" #include "llvm/ADT/SmallSet.h" diff --git a/mlir/lib/IR/ExtensibleDialect.cpp b/mlir/lib/IR/ExtensibleDialect.cpp index ca7b0e1..8a7d747 100644 --- a/mlir/lib/IR/ExtensibleDialect.cpp +++ b/mlir/lib/IR/ExtensibleDialect.cpp @@ -11,7 +11,6 @@ #include "mlir/IR/DialectImplementation.h" #include "mlir/IR/OperationSupport.h" #include "mlir/IR/StorageUniquerSupport.h" -#include "mlir/Support/LogicalResult.h" using namespace mlir; diff --git a/mlir/lib/Query/Query.cpp b/mlir/lib/Query/Query.cpp index 7495e70..7d9f360 100644 --- a/mlir/lib/Query/Query.cpp +++ b/mlir/lib/Query/Query.cpp @@ -12,7 +12,6 @@ #include "mlir/IR/IRMapping.h" #include "mlir/Query/Matcher/MatchFinder.h" #include "mlir/Query/QuerySession.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/SourceMgr.h" #include "llvm/Support/raw_ostream.h" @@ -102,19 +101,16 @@ static Operation *extractFunction(std::vector<Operation *> &ops, Query::~Query() = default; -mlir::LogicalResult InvalidQuery::run(llvm::raw_ostream &os, - QuerySession &qs) const { +LogicalResult InvalidQuery::run(llvm::raw_ostream &os, QuerySession &qs) const { os << errStr << "\n"; return mlir::failure(); } -mlir::LogicalResult NoOpQuery::run(llvm::raw_ostream &os, - QuerySession &qs) const { +LogicalResult NoOpQuery::run(llvm::raw_ostream &os, QuerySession &qs) const { return mlir::success(); } -mlir::LogicalResult HelpQuery::run(llvm::raw_ostream &os, - QuerySession &qs) const { +LogicalResult HelpQuery::run(llvm::raw_ostream &os, QuerySession &qs) const { os << "Available commands:\n\n" " match MATCHER, m MATCHER " "Match the mlir against the given matcher.\n" @@ -123,14 +119,12 @@ mlir::LogicalResult HelpQuery::run(llvm::raw_ostream &os, return mlir::success(); } -mlir::LogicalResult QuitQuery::run(llvm::raw_ostream &os, - QuerySession &qs) const { +LogicalResult QuitQuery::run(llvm::raw_ostream &os, QuerySession &qs) const { qs.terminate = true; return mlir::success(); } -mlir::LogicalResult MatchQuery::run(llvm::raw_ostream &os, - QuerySession &qs) const { +LogicalResult MatchQuery::run(llvm::raw_ostream &os, QuerySession &qs) const { Operation *rootOp = qs.getRootOp(); int matchCount = 0; std::vector<Operation *> matches = diff --git a/mlir/lib/Support/ToolUtilities.cpp b/mlir/lib/Support/ToolUtilities.cpp index f05b4f5..748f928 100644 --- a/mlir/lib/Support/ToolUtilities.cpp +++ b/mlir/lib/Support/ToolUtilities.cpp @@ -12,7 +12,6 @@ #include "mlir/Support/ToolUtilities.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/Support/SourceMgr.h" #include "llvm/Support/raw_ostream.h" diff --git a/mlir/lib/TableGen/CodeGenHelpers.cpp b/mlir/lib/TableGen/CodeGenHelpers.cpp index 5986514..718a813 100644 --- a/mlir/lib/TableGen/CodeGenHelpers.cpp +++ b/mlir/lib/TableGen/CodeGenHelpers.cpp @@ -111,7 +111,7 @@ StringRef StaticVerifierFunctionEmitter::getRegionConstraintFn( /// Code for a type constraint. These may be called on the type of either /// operands or results. static const char *const typeConstraintCode = R"( -static ::mlir::LogicalResult {0}( +static ::llvm::LogicalResult {0}( ::mlir::Operation *op, ::mlir::Type type, ::llvm::StringRef valueKind, unsigned valueIndex) { if (!({1})) { @@ -129,14 +129,14 @@ static ::mlir::LogicalResult {0}( /// TODO: Unique constraints for adaptors. However, most Adaptor::verify /// functions are stripped anyways. static const char *const attrConstraintCode = R"( -static ::mlir::LogicalResult {0}( +static ::llvm::LogicalResult {0}( ::mlir::Attribute attr, ::llvm::StringRef attrName, llvm::function_ref<::mlir::InFlightDiagnostic()> emitError) {{ if (attr && !({1})) return emitError() << "attribute '" << attrName << "' failed to satisfy constraint: {2}"; return ::mlir::success(); } -static ::mlir::LogicalResult {0}( +static ::llvm::LogicalResult {0}( ::mlir::Operation *op, ::mlir::Attribute attr, ::llvm::StringRef attrName) {{ return {0}(attr, attrName, [op]() {{ return op->emitOpError(); @@ -146,7 +146,7 @@ static ::mlir::LogicalResult {0}( /// Code for a successor constraint. static const char *const successorConstraintCode = R"( -static ::mlir::LogicalResult {0}( +static ::llvm::LogicalResult {0}( ::mlir::Operation *op, ::mlir::Block *successor, ::llvm::StringRef successorName, unsigned successorIndex) { if (!({1})) { @@ -160,7 +160,7 @@ static ::mlir::LogicalResult {0}( /// Code for a region constraint. Callers will need to pass in the region's name /// for emitting an error message. static const char *const regionConstraintCode = R"( -static ::mlir::LogicalResult {0}( +static ::llvm::LogicalResult {0}( ::mlir::Operation *op, ::mlir::Region ®ion, ::llvm::StringRef regionName, unsigned regionIndex) { if (!({1})) { @@ -176,7 +176,7 @@ static ::mlir::LogicalResult {0}( /// /// {3}: "Type type" or "Attribute attr". static const char *const patternAttrOrTypeConstraintCode = R"( -static ::mlir::LogicalResult {0}( +static ::llvm::LogicalResult {0}( ::mlir::PatternRewriter &rewriter, ::mlir::Operation *op, ::mlir::{3}, ::llvm::StringRef failureStr) { if (!({1})) { diff --git a/mlir/lib/Target/LLVM/NVVM/Target.cpp b/mlir/lib/Target/LLVM/NVVM/Target.cpp index acb903a..e608d26 100644 --- a/mlir/lib/Target/LLVM/NVVM/Target.cpp +++ b/mlir/lib/Target/LLVM/NVVM/Target.cpp @@ -157,7 +157,6 @@ SerializeGPUModuleBase::loadBitcodeFiles(llvm::Module &module) { return std::move(bcFiles); } -#if LLVM_HAS_NVPTX_TARGET namespace { class NVPTXSerializer : public SerializeGPUModuleBase { public: @@ -532,6 +531,12 @@ NVPTXSerializer::moduleToObject(llvm::Module &llvmModule) { if (targetOptions.getCompilationTarget() == gpu::CompilationTarget::Offload) return SerializeGPUModuleBase::moduleToObject(llvmModule); +#if !LLVM_HAS_NVPTX_TARGET + getOperation()->emitError( + "The `NVPTX` target was not built. Please enable it when building LLVM."); + return std::nullopt; +#endif // LLVM_HAS_NVPTX_TARGET + // Emit PTX code. std::optional<llvm::TargetMachine *> targetMachine = getOrCreateTargetMachine(); @@ -569,7 +574,6 @@ NVPTXSerializer::moduleToObject(llvm::Module &llvmModule) { return compileToBinary(*serializedISA); #endif // MLIR_ENABLE_NVPTXCOMPILER } -#endif // LLVM_HAS_NVPTX_TARGET std::optional<SmallVector<char, 0>> NVVMTargetAttrImpl::serializeToObject(Attribute attribute, Operation *module, @@ -581,15 +585,9 @@ NVVMTargetAttrImpl::serializeToObject(Attribute attribute, Operation *module, module->emitError("Module must be a GPU module."); return std::nullopt; } -#if LLVM_HAS_NVPTX_TARGET NVPTXSerializer serializer(*module, cast<NVVMTargetAttr>(attribute), options); serializer.init(); return serializer.run(); -#else - module->emitError( - "The `NVPTX` target was not built. Please enable it when building LLVM."); - return std::nullopt; -#endif // LLVM_HAS_NVPTX_TARGET } Attribute diff --git a/mlir/lib/Target/LLVMIR/Dialect/GPU/SelectObjectAttr.cpp b/mlir/lib/Target/LLVMIR/Dialect/GPU/SelectObjectAttr.cpp index 0eb3328..b023c4c 100644 --- a/mlir/lib/Target/LLVMIR/Dialect/GPU/SelectObjectAttr.cpp +++ b/mlir/lib/Target/LLVMIR/Dialect/GPU/SelectObjectAttr.cpp @@ -167,7 +167,7 @@ public: Value *createKernelArgArray(mlir::gpu::LaunchFuncOp op); // Create the full kernel launch. - mlir::LogicalResult createKernelLaunch(mlir::gpu::LaunchFuncOp op, + llvm::LogicalResult createKernelLaunch(mlir::gpu::LaunchFuncOp op, mlir::gpu::ObjectAttr object); private: @@ -345,7 +345,7 @@ llvm::LaunchKernel::createKernelArgArray(mlir::gpu::LaunchFuncOp op) { // call %streamSynchronize(%4) // call %streamDestroy(%4) // call %moduleUnload(%1) -mlir::LogicalResult +llvm::LogicalResult llvm::LaunchKernel::createKernelLaunch(mlir::gpu::LaunchFuncOp op, mlir::gpu::ObjectAttr object) { auto llvmValue = [&](mlir::Value value) -> Value * { diff --git a/mlir/lib/Target/LLVMIR/Dialect/NVVM/NVVMToLLVMIRTranslation.cpp b/mlir/lib/Target/LLVMIR/Dialect/NVVM/NVVMToLLVMIRTranslation.cpp index ea9fe26..a09c24d 100644 --- a/mlir/lib/Target/LLVMIR/Dialect/NVVM/NVVMToLLVMIRTranslation.cpp +++ b/mlir/lib/Target/LLVMIR/Dialect/NVVM/NVVMToLLVMIRTranslation.cpp @@ -15,7 +15,6 @@ #include "mlir/Dialect/LLVMIR/NVVMDialect.h" #include "mlir/Dialect/Utils/StaticValueUtils.h" #include "mlir/IR/Operation.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Target/LLVMIR/ModuleTranslation.h" #include "llvm/IR/IRBuilder.h" diff --git a/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp b/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp index fe257a4..35971fb 100644 --- a/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp +++ b/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp @@ -18,7 +18,6 @@ #include "mlir/IR/IRMapping.h" #include "mlir/IR/Operation.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Target/LLVMIR/Dialect/OpenMPCommon.h" #include "mlir/Target/LLVMIR/ModuleTranslation.h" #include "mlir/Transforms/RegionUtils.h" diff --git a/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp b/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp index 6e8b2de..e0b1816e 100644 --- a/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp +++ b/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp @@ -31,7 +31,6 @@ #include "mlir/IR/DialectResourceBlobManager.h" #include "mlir/IR/RegionGraphTraits.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Target/LLVMIR/LLVMTranslationInterface.h" #include "mlir/Target/LLVMIR/TypeToLLVM.h" diff --git a/mlir/lib/Target/SPIRV/Deserialization/Deserializer.cpp b/mlir/lib/Target/SPIRV/Deserialization/Deserializer.cpp index faaa420..d7a3085 100644 --- a/mlir/lib/Target/SPIRV/Deserialization/Deserializer.cpp +++ b/mlir/lib/Target/SPIRV/Deserialization/Deserializer.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/Builders.h" #include "mlir/IR/IRMapping.h" #include "mlir/IR/Location.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Target/SPIRV/SPIRVBinaryUtils.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/Sequence.h" diff --git a/mlir/lib/Target/SPIRV/Serialization/SerializeOps.cpp b/mlir/lib/Target/SPIRV/Serialization/SerializeOps.cpp index c283e64..6c28c04 100644 --- a/mlir/lib/Target/SPIRV/Serialization/SerializeOps.cpp +++ b/mlir/lib/Target/SPIRV/Serialization/SerializeOps.cpp @@ -15,7 +15,6 @@ #include "mlir/Dialect/SPIRV/IR/SPIRVAttributes.h" #include "mlir/Dialect/SPIRV/IR/SPIRVEnums.h" #include "mlir/IR/RegionGraphTraits.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Target/SPIRV/SPIRVBinaryUtils.h" #include "llvm/ADT/DepthFirstIterator.h" #include "llvm/ADT/StringExtras.h" diff --git a/mlir/lib/Target/SPIRV/Serialization/Serializer.cpp b/mlir/lib/Target/SPIRV/Serialization/Serializer.cpp index 200abdf..4c4fef1 100644 --- a/mlir/lib/Target/SPIRV/Serialization/Serializer.cpp +++ b/mlir/lib/Target/SPIRV/Serialization/Serializer.cpp @@ -16,7 +16,6 @@ #include "mlir/Dialect/SPIRV/IR/SPIRVDialect.h" #include "mlir/Dialect/SPIRV/IR/SPIRVEnums.h" #include "mlir/Dialect/SPIRV/IR/SPIRVTypes.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Target/SPIRV/SPIRVBinaryUtils.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/Sequence.h" diff --git a/mlir/lib/Tools/PDLL/CodeGen/CPPGen.cpp b/mlir/lib/Tools/PDLL/CodeGen/CPPGen.cpp index b02edc6..611b734 100644 --- a/mlir/lib/Tools/PDLL/CodeGen/CPPGen.cpp +++ b/mlir/lib/Tools/PDLL/CodeGen/CPPGen.cpp @@ -216,7 +216,7 @@ void CodeGen::generateConstraintOrRewrite(const ast::CallableDecl *decl, // TODO: This will need to change if we allow Constraints to return values as // well. if (isConstraint) { - os << "::mlir::LogicalResult"; + os << "::llvm::LogicalResult"; } else { // Otherwise, generate a type based on the results of the callable. // If the callable has explicit results, use those to build the result. diff --git a/mlir/lib/Tools/PDLL/Parser/Lexer.cpp b/mlir/lib/Tools/PDLL/Parser/Lexer.cpp index 4673a73b..cc111a0 100644 --- a/mlir/lib/Tools/PDLL/Parser/Lexer.cpp +++ b/mlir/lib/Tools/PDLL/Parser/Lexer.cpp @@ -7,7 +7,6 @@ //===----------------------------------------------------------------------===// #include "Lexer.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Tools/PDLL/AST/Diagnostic.h" #include "mlir/Tools/PDLL/Parser/CodeComplete.h" #include "llvm/ADT/StringExtras.h" diff --git a/mlir/lib/Tools/PDLL/Parser/Lexer.h b/mlir/lib/Tools/PDLL/Parser/Lexer.h index c80cb36..cd9c2b7 100644 --- a/mlir/lib/Tools/PDLL/Parser/Lexer.h +++ b/mlir/lib/Tools/PDLL/Parser/Lexer.h @@ -18,8 +18,6 @@ class SourceMgr; } // namespace llvm namespace mlir { -struct LogicalResult; - namespace pdll { class CodeCompleteContext; diff --git a/mlir/lib/Tools/PDLL/Parser/Parser.cpp b/mlir/lib/Tools/PDLL/Parser/Parser.cpp index 45f9e2f..1f0df03 100644 --- a/mlir/lib/Tools/PDLL/Parser/Parser.cpp +++ b/mlir/lib/Tools/PDLL/Parser/Parser.cpp @@ -9,7 +9,6 @@ #include "mlir/Tools/PDLL/Parser/Parser.h" #include "Lexer.h" #include "mlir/Support/IndentedOstream.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/TableGen/Argument.h" #include "mlir/TableGen/Attribute.h" #include "mlir/TableGen/Constraint.h" diff --git a/mlir/lib/Tools/lsp-server-support/Protocol.cpp b/mlir/lib/Tools/lsp-server-support/Protocol.cpp index 188f525..e4eb251 100644 --- a/mlir/lib/Tools/lsp-server-support/Protocol.cpp +++ b/mlir/lib/Tools/lsp-server-support/Protocol.cpp @@ -11,7 +11,6 @@ //===----------------------------------------------------------------------===// #include "mlir/Tools/lsp-server-support/Protocol.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Tools/lsp-server-support/Logging.h" #include "llvm/ADT/Hashing.h" #include "llvm/ADT/SmallString.h" diff --git a/mlir/lib/Tools/mlir-lsp-server/LSPServer.h b/mlir/lib/Tools/mlir-lsp-server/LSPServer.h index faf5936..2c50c6b 100644 --- a/mlir/lib/Tools/mlir-lsp-server/LSPServer.h +++ b/mlir/lib/Tools/mlir-lsp-server/LSPServer.h @@ -11,16 +11,19 @@ #include <memory> -namespace mlir { +namespace llvm { struct LogicalResult; +} // namespace llvm +namespace mlir { namespace lsp { class JSONTransport; class MLIRServer; /// Run the main loop of the LSP server using the given MLIR server and /// transport. -LogicalResult runMlirLSPServer(MLIRServer &server, JSONTransport &transport); +llvm::LogicalResult runMlirLSPServer(MLIRServer &server, + JSONTransport &transport); } // namespace lsp } // namespace mlir diff --git a/mlir/lib/Tools/mlir-opt/MlirOptMain.cpp b/mlir/lib/Tools/mlir-opt/MlirOptMain.cpp index a1b2893..831e1e6 100644 --- a/mlir/lib/Tools/mlir-opt/MlirOptMain.cpp +++ b/mlir/lib/Tools/mlir-opt/MlirOptMain.cpp @@ -31,7 +31,6 @@ #include "mlir/Pass/Pass.h" #include "mlir/Pass/PassManager.h" #include "mlir/Support/FileUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/Timing.h" #include "mlir/Support/ToolUtilities.h" #include "mlir/Tools/ParseUtilities.h" diff --git a/mlir/lib/Tools/mlir-pdll-lsp-server/LSPServer.h b/mlir/lib/Tools/mlir-pdll-lsp-server/LSPServer.h index 5551c4e..78c4c31 100644 --- a/mlir/lib/Tools/mlir-pdll-lsp-server/LSPServer.h +++ b/mlir/lib/Tools/mlir-pdll-lsp-server/LSPServer.h @@ -11,16 +11,19 @@ #include <memory> -namespace mlir { +namespace llvm { struct LogicalResult; +} // namespace llvm +namespace mlir { namespace lsp { class JSONTransport; class PDLLServer; /// Run the main loop of the LSP server using the given PDLL server and /// transport. -LogicalResult runPdllLSPServer(PDLLServer &server, JSONTransport &transport); +llvm::LogicalResult runPdllLSPServer(PDLLServer &server, + JSONTransport &transport); } // namespace lsp } // namespace mlir diff --git a/mlir/lib/Tools/mlir-pdll-lsp-server/PDLLServer.h b/mlir/lib/Tools/mlir-pdll-lsp-server/PDLLServer.h index 682c7b5..134431f 100644 --- a/mlir/lib/Tools/mlir-pdll-lsp-server/PDLLServer.h +++ b/mlir/lib/Tools/mlir-pdll-lsp-server/PDLLServer.h @@ -12,8 +12,9 @@ #include "mlir/Support/LLVM.h" #include "llvm/ADT/StringRef.h" #include <memory> -#include <string> #include <optional> +#include <string> +#include <vector> namespace mlir { namespace lsp { diff --git a/mlir/lib/Tools/mlir-query/MlirQueryMain.cpp b/mlir/lib/Tools/mlir-query/MlirQueryMain.cpp index 15de16a..9950050 100644 --- a/mlir/lib/Tools/mlir-query/MlirQueryMain.cpp +++ b/mlir/lib/Tools/mlir-query/MlirQueryMain.cpp @@ -18,7 +18,6 @@ #include "mlir/Query/Query.h" #include "mlir/Query/QuerySession.h" #include "mlir/Support/FileUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/LineEditor/LineEditor.h" #include "llvm/Support/CommandLine.h" #include "llvm/Support/InitLLVM.h" @@ -28,7 +27,7 @@ // Query Parser //===----------------------------------------------------------------------===// -mlir::LogicalResult +llvm::LogicalResult mlir::mlirQueryMain(int argc, char **argv, MLIRContext &context, const mlir::query::matcher::Registry &matcherRegistry) { diff --git a/mlir/lib/Tools/mlir-reduce/MlirReduceMain.cpp b/mlir/lib/Tools/mlir-reduce/MlirReduceMain.cpp index 2f0ab8a..a9cde4a 100644 --- a/mlir/lib/Tools/mlir-reduce/MlirReduceMain.cpp +++ b/mlir/lib/Tools/mlir-reduce/MlirReduceMain.cpp @@ -21,7 +21,6 @@ #include "mlir/Reducer/Passes.h" #include "mlir/Rewrite/FrozenRewritePatternSet.h" #include "mlir/Support/FileUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Tools/ParseUtilities.h" #include "llvm/Support/InitLLVM.h" #include "llvm/Support/SourceMgr.h" diff --git a/mlir/lib/Tools/mlir-translate/MlirTranslateMain.cpp b/mlir/lib/Tools/mlir-translate/MlirTranslateMain.cpp index bd992895..56773f5 100644 --- a/mlir/lib/Tools/mlir-translate/MlirTranslateMain.cpp +++ b/mlir/lib/Tools/mlir-translate/MlirTranslateMain.cpp @@ -13,7 +13,6 @@ #include "mlir/IR/Verifier.h" #include "mlir/Parser/Parser.h" #include "mlir/Support/FileUtilities.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Support/Timing.h" #include "mlir/Support/ToolUtilities.h" #include "mlir/Tools/mlir-translate/Translation.h" diff --git a/mlir/lib/Tools/tblgen-lsp-server/LSPServer.h b/mlir/lib/Tools/tblgen-lsp-server/LSPServer.h index bc85de1..501a9da 100644 --- a/mlir/lib/Tools/tblgen-lsp-server/LSPServer.h +++ b/mlir/lib/Tools/tblgen-lsp-server/LSPServer.h @@ -11,17 +11,19 @@ #include <memory> -namespace mlir { +namespace llvm { struct LogicalResult; +} // namespace llvm +namespace mlir { namespace lsp { class JSONTransport; class TableGenServer; /// Run the main loop of the LSP server using the given TableGen server and /// transport. -LogicalResult runTableGenLSPServer(TableGenServer &server, - JSONTransport &transport); +llvm::LogicalResult runTableGenLSPServer(TableGenServer &server, + JSONTransport &transport); } // namespace lsp } // namespace mlir diff --git a/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.cpp b/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.cpp index 5ea91a2..f7cf4de 100644 --- a/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.cpp +++ b/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.cpp @@ -9,7 +9,6 @@ #include "TableGenServer.h" #include "mlir/Support/IndentedOstream.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Tools/lsp-server-support/CompilationDatabase.h" #include "mlir/Tools/lsp-server-support/Logging.h" #include "mlir/Tools/lsp-server-support/Protocol.h" @@ -17,9 +16,6 @@ #include "llvm/ADT/IntervalMap.h" #include "llvm/ADT/PointerUnion.h" #include "llvm/ADT/StringMap.h" -#include "llvm/ADT/StringSet.h" -#include "llvm/ADT/TypeSwitch.h" -#include "llvm/Support/FileSystem.h" #include "llvm/Support/Path.h" #include "llvm/TableGen/Parser.h" #include "llvm/TableGen/Record.h" diff --git a/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.h b/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.h index d749a4f..bdc8510 100644 --- a/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.h +++ b/mlir/lib/Tools/tblgen-lsp-server/TableGenServer.h @@ -12,8 +12,9 @@ #include "mlir/Support/LLVM.h" #include "llvm/ADT/StringRef.h" #include <memory> -#include <string> #include <optional> +#include <string> +#include <vector> namespace mlir { namespace lsp { diff --git a/mlir/test/lib/Analysis/DataFlow/TestLivenessAnalysis.cpp b/mlir/test/lib/Analysis/DataFlow/TestLivenessAnalysis.cpp index 6a67ddd..43005e2 100644 --- a/mlir/test/lib/Analysis/DataFlow/TestLivenessAnalysis.cpp +++ b/mlir/test/lib/Analysis/DataFlow/TestLivenessAnalysis.cpp @@ -18,7 +18,6 @@ #include <mlir/Pass/Pass.h> #include <mlir/Pass/PassRegistry.h> #include <mlir/Support/LLVM.h> -#include <mlir/Support/LogicalResult.h> #include <mlir/Support/TypeID.h> using namespace mlir; diff --git a/mlir/test/lib/Dialect/ArmNeon/TestLowerToArmNeon.cpp b/mlir/test/lib/Dialect/ArmNeon/TestLowerToArmNeon.cpp index 2e5a4a9..f6bfd9f 100644 --- a/mlir/test/lib/Dialect/ArmNeon/TestLowerToArmNeon.cpp +++ b/mlir/test/lib/Dialect/ArmNeon/TestLowerToArmNeon.cpp @@ -17,7 +17,6 @@ #include "mlir/IR/PatternMatch.h" #include "mlir/Pass/Pass.h" #include "mlir/Pass/PassManager.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" #define PASS_NAME "test-lower-to-arm-neon" diff --git a/mlir/test/lib/Dialect/Mesh/TestOpLowering.cpp b/mlir/test/lib/Dialect/Mesh/TestOpLowering.cpp index 321b6a4..1f836be 100644 --- a/mlir/test/lib/Dialect/Mesh/TestOpLowering.cpp +++ b/mlir/test/lib/Dialect/Mesh/TestOpLowering.cpp @@ -11,7 +11,6 @@ #include "mlir/Dialect/Utils/IndexingUtils.h" #include "mlir/IR/SymbolTable.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" using namespace mlir; diff --git a/mlir/test/lib/Dialect/Mesh/TestReshardingSpmdization.cpp b/mlir/test/lib/Dialect/Mesh/TestReshardingSpmdization.cpp index 5e3918f..f964102 100644 --- a/mlir/test/lib/Dialect/Mesh/TestReshardingSpmdization.cpp +++ b/mlir/test/lib/Dialect/Mesh/TestReshardingSpmdization.cpp @@ -19,7 +19,6 @@ #include "mlir/IR/SymbolTable.h" #include "mlir/IR/Value.h" #include "mlir/Pass/Pass.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/GreedyPatternRewriteDriver.h" using namespace mlir; diff --git a/mlir/test/lib/Dialect/Test/TestAttributes.cpp b/mlir/test/lib/Dialect/Test/TestAttributes.cpp index d7e40d3..b66dfbf 100644 --- a/mlir/test/lib/Dialect/Test/TestAttributes.cpp +++ b/mlir/test/lib/Dialect/Test/TestAttributes.cpp @@ -17,7 +17,6 @@ #include "mlir/IR/DialectImplementation.h" #include "mlir/IR/ExtensibleDialect.h" #include "mlir/IR/Types.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/APFloat.h" #include "llvm/ADT/Hashing.h" #include "llvm/ADT/StringExtras.h" diff --git a/mlir/test/lib/Dialect/Test/TestDialect.cpp b/mlir/test/lib/Dialect/Test/TestDialect.cpp index bfb9592..1bbf2cc 100644 --- a/mlir/test/lib/Dialect/Test/TestDialect.cpp +++ b/mlir/test/lib/Dialect/Test/TestDialect.cpp @@ -28,7 +28,6 @@ #include "mlir/Interfaces/FunctionImplementation.h" #include "mlir/Interfaces/InferIntRangeInterface.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/Transforms/FoldUtils.h" #include "mlir/Transforms/InliningUtils.h" #include "llvm/ADT/STLFunctionalExtras.h" diff --git a/mlir/test/lib/Dialect/Test/TestOpDefs.cpp b/mlir/test/lib/Dialect/Test/TestOpDefs.cpp index b86f692..fbaa102 100644 --- a/mlir/test/lib/Dialect/Test/TestOpDefs.cpp +++ b/mlir/test/lib/Dialect/Test/TestOpDefs.cpp @@ -1193,7 +1193,7 @@ void TestVersionedOpA::writeProperties(mlir::DialectBytecodeWriter &writer) { // TestOpWithVersionedProperties //===----------------------------------------------------------------------===// -mlir::LogicalResult TestOpWithVersionedProperties::readFromMlirBytecode( +llvm::LogicalResult TestOpWithVersionedProperties::readFromMlirBytecode( mlir::DialectBytecodeReader &reader, test::VersionedProperties &prop) { uint64_t value1, value2 = 0; if (failed(reader.readVarInt(value1))) diff --git a/mlir/test/lib/Dialect/Test/TestOps.h b/mlir/test/lib/Dialect/Test/TestOps.h index 837ccca..f070c3b 100644 --- a/mlir/test/lib/Dialect/Test/TestOps.h +++ b/mlir/test/lib/Dialect/Test/TestOps.h @@ -70,7 +70,7 @@ struct PropertiesWithCustomPrint { } }; -mlir::LogicalResult setPropertiesFromAttribute( +llvm::LogicalResult setPropertiesFromAttribute( PropertiesWithCustomPrint &prop, mlir::Attribute attr, llvm::function_ref<mlir::InFlightDiagnostic()> emitError); mlir::DictionaryAttr @@ -92,7 +92,7 @@ public: // These three methods are invoked through the `MyStructProperty` wrapper // defined in TestOps.td mlir::Attribute asAttribute(mlir::MLIRContext *ctx) const; - static mlir::LogicalResult + static llvm::LogicalResult setFromAttr(MyPropStruct &prop, mlir::Attribute attr, llvm::function_ref<mlir::InFlightDiagnostic()> emitError); llvm::hash_code hash() const; @@ -101,7 +101,7 @@ public: } }; -mlir::LogicalResult readFromMlirBytecode(mlir::DialectBytecodeReader &reader, +llvm::LogicalResult readFromMlirBytecode(mlir::DialectBytecodeReader &reader, MyPropStruct &prop); void writeToMlirBytecode(mlir::DialectBytecodeWriter &writer, MyPropStruct &prop); @@ -122,7 +122,7 @@ struct VersionedProperties { } }; -mlir::LogicalResult setPropertiesFromAttribute( +llvm::LogicalResult setPropertiesFromAttribute( VersionedProperties &prop, mlir::Attribute attr, llvm::function_ref<mlir::InFlightDiagnostic()> emitError); mlir::DictionaryAttr getPropertiesAsAttribute(mlir::MLIRContext *ctx, @@ -137,7 +137,7 @@ mlir::ParseResult customParseProperties(mlir::OpAsmParser &parser, // Bytecode Support //===----------------------------------------------------------------------===// -mlir::LogicalResult readFromMlirBytecode(mlir::DialectBytecodeReader &reader, +llvm::LogicalResult readFromMlirBytecode(mlir::DialectBytecodeReader &reader, llvm::MutableArrayRef<int64_t> prop); void writeToMlirBytecode(mlir::DialectBytecodeWriter &writer, llvm::ArrayRef<int64_t> prop); diff --git a/mlir/test/lib/Dialect/Test/TestOps.td b/mlir/test/lib/Dialect/Test/TestOps.td index c77f76d..e1ec142 100644 --- a/mlir/test/lib/Dialect/Test/TestOps.td +++ b/mlir/test/lib/Dialect/Test/TestOps.td @@ -462,7 +462,7 @@ def VariadicRegionInferredTypesOp : TEST_Op<"variadic_region_inferred", let results = (outs Variadic<AnyType>); let extraClassDeclaration = [{ - static mlir::LogicalResult inferReturnTypes(mlir::MLIRContext *context, + static llvm::LogicalResult inferReturnTypes(mlir::MLIRContext *context, std::optional<::mlir::Location> location, mlir::ValueRange operands, mlir::DictionaryAttr attributes, mlir::OpaqueProperties properties, mlir::RegionRange regions, llvm::SmallVectorImpl<mlir::Type> &inferredReturnTypes) { @@ -1368,7 +1368,7 @@ def TestOpInPlaceFoldSuccess : TEST_Op<"op_in_place_fold_success"> { let results = (outs Variadic<I1>); let hasFolder = 1; let extraClassDefinition = [{ - ::mlir::LogicalResult $cppClass::fold(FoldAdaptor adaptor, + ::llvm::LogicalResult $cppClass::fold(FoldAdaptor adaptor, SmallVectorImpl<OpFoldResult> &results) { return success(); } @@ -2408,7 +2408,7 @@ class TableGenBuildInferReturnTypeBaseOp<string mnemonic, let results = (outs AnyType:$result); let extraClassDeclaration = [{ - static ::mlir::LogicalResult inferReturnTypes(::mlir::MLIRContext *, + static ::llvm::LogicalResult inferReturnTypes(::mlir::MLIRContext *, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, mlir::OpaqueProperties properties, ::mlir::RegionRange regions, ::llvm::SmallVectorImpl<::mlir::Type> &inferredReturnTypes) { @@ -3037,7 +3037,7 @@ def TestOpWithNiceProperties : TEST_Op<"with_nice_properties"> { ::mlir::ArrayRef<::llvm::StringRef> elidedProps); static ::mlir::ParseResult parseProperties(::mlir::OpAsmParser &parser, ::mlir::OperationState &result); - static ::mlir::LogicalResult readFromMlirBytecode( + static ::llvm::LogicalResult readFromMlirBytecode( ::mlir::DialectBytecodeReader &, test::PropertiesWithCustomPrint &prop); static void writeToMlirBytecode( @@ -3045,7 +3045,7 @@ def TestOpWithNiceProperties : TEST_Op<"with_nice_properties"> { const test::PropertiesWithCustomPrint &prop); }]; let extraClassDefinition = [{ - ::mlir::LogicalResult TestOpWithNiceProperties::readFromMlirBytecode( + ::llvm::LogicalResult TestOpWithNiceProperties::readFromMlirBytecode( ::mlir::DialectBytecodeReader &reader, test::PropertiesWithCustomPrint &prop) { StringRef label; @@ -3101,7 +3101,7 @@ def TestOpWithVersionedProperties : TEST_Op<"with_versioned_properties"> { ::mlir::ArrayRef<::llvm::StringRef> elidedProps); static ::mlir::ParseResult parseProperties(::mlir::OpAsmParser &parser, ::mlir::OperationState &result); - static ::mlir::LogicalResult readFromMlirBytecode( + static ::llvm::LogicalResult readFromMlirBytecode( ::mlir::DialectBytecodeReader &, test::VersionedProperties &prop); static void writeToMlirBytecode( diff --git a/mlir/test/lib/Dialect/Test/TestOpsSyntax.cpp b/mlir/test/lib/Dialect/Test/TestOpsSyntax.cpp index ebaced5..664951f 100644 --- a/mlir/test/lib/Dialect/Test/TestOpsSyntax.cpp +++ b/mlir/test/lib/Dialect/Test/TestOpsSyntax.cpp @@ -280,7 +280,7 @@ void ParseB64BytesOp::print(OpAsmPrinter &p) { p << " \"" << llvm::encodeBase64(getB64()) << "\""; } -::mlir::LogicalResult FormatInferType2Op::inferReturnTypes( +::llvm::LogicalResult FormatInferType2Op::inferReturnTypes( ::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, OpaqueProperties properties, ::mlir::RegionRange regions, diff --git a/mlir/test/lib/Dialect/Test/TestOpsSyntax.td b/mlir/test/lib/Dialect/Test/TestOpsSyntax.td index 9522a77..3129085 100644 --- a/mlir/test/lib/Dialect/Test/TestOpsSyntax.td +++ b/mlir/test/lib/Dialect/Test/TestOpsSyntax.td @@ -632,7 +632,7 @@ def FormatInferTypeOp : TEST_Op<"format_infer_type", [InferTypeOpInterface]> { let assemblyFormat = "attr-dict"; let extraClassDeclaration = [{ - static ::mlir::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, + static ::llvm::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, mlir::OpaqueProperties properties, ::mlir::RegionRange regions, ::llvm::SmallVectorImpl<::mlir::Type> &inferredReturnTypes) { @@ -655,7 +655,7 @@ class FormatInferAllTypesBaseOp<string mnemonic, list<Trait> traits = []> let arguments = (ins Variadic<AnyType>:$args); let results = (outs Variadic<AnyType>:$outs); let extraClassDeclaration = [{ - static ::mlir::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, + static ::llvm::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, mlir::OpaqueProperties properties, ::mlir::RegionRange regions, ::llvm::SmallVectorImpl<::mlir::Type> &inferredReturnTypes) { @@ -702,7 +702,7 @@ def FormatInferTypeRegionsOp let regions = (region AnyRegion:$region); let assemblyFormat = "$region attr-dict"; let extraClassDeclaration = [{ - static ::mlir::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, + static ::llvm::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, mlir::OpaqueProperties properties, ::mlir::RegionRange regions, ::llvm::SmallVectorImpl<::mlir::Type> &inferredReturnTypes) { @@ -723,7 +723,7 @@ def FormatInferTypeVariadicOperandsOp let results = (outs Variadic<AnyType>:$outs); let assemblyFormat = "`(` $a `:` type($a) `)` `(` $b `:` type($b) `)` attr-dict"; let extraClassDeclaration = [{ - static ::mlir::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, + static ::llvm::LogicalResult inferReturnTypes(::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, mlir::OpaqueProperties properties, ::mlir::RegionRange regions, ::llvm::SmallVectorImpl<::mlir::Type> &inferredReturnTypes) { diff --git a/mlir/test/lib/Dialect/Test/TestTypeDefs.td b/mlir/test/lib/Dialect/Test/TestTypeDefs.td index 492642b..d96152a 100644 --- a/mlir/test/lib/Dialect/Test/TestTypeDefs.td +++ b/mlir/test/lib/Dialect/Test/TestTypeDefs.td @@ -154,7 +154,7 @@ def TestTypeWithLayoutType : Test_Type<"TestTypeWithLayout", [ let mnemonic = "test_type_with_layout"; let parameters = (ins "unsigned":$key); let extraClassDeclaration = [{ - ::mlir::LogicalResult verifyEntries(::mlir::DataLayoutEntryListRef params, + ::llvm::LogicalResult verifyEntries(::mlir::DataLayoutEntryListRef params, ::mlir::Location loc) const; private: diff --git a/mlir/test/lib/Dialect/Test/TestTypes.h b/mlir/test/lib/Dialect/Test/TestTypes.h index da56049..cef3f05 100644 --- a/mlir/test/lib/Dialect/Test/TestTypes.h +++ b/mlir/test/lib/Dialect/Test/TestTypes.h @@ -108,7 +108,7 @@ struct TestRecursiveTypeStorage : public ::mlir::TypeStorage { TestRecursiveTypeStorage(allocator.copyInto(key)); } - ::mlir::LogicalResult mutate(::mlir::TypeStorageAllocator &allocator, + ::llvm::LogicalResult mutate(::mlir::TypeStorageAllocator &allocator, ::mlir::Type newBody) { // Cannot set a different body than before. if (body && body != newBody) @@ -140,7 +140,7 @@ public: } /// Body getter and setter. - ::mlir::LogicalResult setBody(Type body) { return Base::mutate(body); } + ::llvm::LogicalResult setBody(Type body) { return Base::mutate(body); } ::mlir::Type getBody() const { return getImpl()->body; } /// Name/key getter. diff --git a/mlir/test/mlir-pdll/CodeGen/CPP/general.pdll b/mlir/test/mlir-pdll/CodeGen/CPP/general.pdll index 21a8966..4e869e5 100644 --- a/mlir/test/mlir-pdll/CodeGen/CPP/general.pdll +++ b/mlir/test/mlir-pdll/CodeGen/CPP/general.pdll @@ -46,7 +46,7 @@ Pattern => erase op<test.op3>; #include "include/ods.td" -// CHECK: static ::mlir::LogicalResult TestCstPDLFn(::mlir::PatternRewriter &rewriter, +// CHECK: static ::llvm::LogicalResult TestCstPDLFn(::mlir::PatternRewriter &rewriter, // CHECK-SAME: ::mlir::Attribute attr, ::mlir::Operation * op, ::mlir::Type type, // CHECK-SAME: ::mlir::Value value, ::mlir::TypeRange typeRange, ::mlir::ValueRange valueRange) { // CHECK-NEXT: return success(); diff --git a/mlir/test/mlir-tblgen/attrdefs.td b/mlir/test/mlir-tblgen/attrdefs.td index 5683f34..35d2c49 100644 --- a/mlir/test/mlir-tblgen/attrdefs.td +++ b/mlir/test/mlir-tblgen/attrdefs.td @@ -86,7 +86,7 @@ def B_CompoundAttrA : TestAttr<"CompoundA"> { // DECL-LABEL: class CompoundAAttr : public ::mlir::Attribute // DECL: static CompoundAAttr getChecked(::llvm::function_ref<::mlir::InFlightDiagnostic()> emitError, ::mlir::MLIRContext *context, int widthOfSomething, ::test::SimpleTypeA exampleTdType, ::llvm::APFloat apFloat, ::llvm::ArrayRef<int> dims, ::mlir::Type inner); -// DECL: static ::mlir::LogicalResult verify(::llvm::function_ref<::mlir::InFlightDiagnostic()> emitError, int widthOfSomething, ::test::SimpleTypeA exampleTdType, ::llvm::APFloat apFloat, ::llvm::ArrayRef<int> dims, ::mlir::Type inner); +// DECL: static ::llvm::LogicalResult verify(::llvm::function_ref<::mlir::InFlightDiagnostic()> emitError, int widthOfSomething, ::test::SimpleTypeA exampleTdType, ::llvm::APFloat apFloat, ::llvm::ArrayRef<int> dims, ::mlir::Type inner); // DECL: static constexpr ::llvm::StringLiteral getMnemonic() { // DECL: return {"cmpnd_a"}; // DECL: } diff --git a/mlir/test/mlir-tblgen/constraint-unique.td b/mlir/test/mlir-tblgen/constraint-unique.td index cd41efa..214cd6c 100644 --- a/mlir/test/mlir-tblgen/constraint-unique.td +++ b/mlir/test/mlir-tblgen/constraint-unique.td @@ -54,7 +54,7 @@ def OpC : NS_Op<"op_c"> { } /// Test that a type contraint was generated. -// CHECK: static ::mlir::LogicalResult [[$A_TYPE_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$A_TYPE_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( // CHECK: if (!((typePred(type, *op)))) { // CHECK-NEXT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NEXT: << " must be a type, but got " << type; @@ -63,13 +63,13 @@ def OpC : NS_Op<"op_c"> { // CHECK-NOT: << " must be a type, but got " << type; /// Test that a type constraint with a different description was generated. -// CHECK: static ::mlir::LogicalResult [[$O_TYPE_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$O_TYPE_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( // CHECK: if (!((typePred(type, *op)))) { // CHECK-NEXT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NEXT: << " must be another type, but got " << type; /// Test that an attribute contraint was generated. -// CHECK: static ::mlir::LogicalResult [[$A_ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$A_ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( // CHECK: if (attr && !((attrPred(attr, *op)))) // CHECK-NEXT: return emitError() << "attribute '" << attrName // CHECK-NEXT: << "' failed to satisfy constraint: an attribute"; @@ -78,14 +78,14 @@ def OpC : NS_Op<"op_c"> { // CHECK-NOT: << "' failed to satisfy constraint: an attribute"; /// Test that a attribute constraint with a different description was generated. -// CHECK: static ::mlir::LogicalResult [[$O_ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( -// CHECK: static ::mlir::LogicalResult [[$O_ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$O_ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$O_ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( // CHECK: if (attr && !((attrPred(attr, *op)))) // CHECK-NEXT: return emitError() << "attribute '" << attrName // CHECK-NEXT: << "' failed to satisfy constraint: another attribute"; /// Test that a successor contraint was generated. -// CHECK: static ::mlir::LogicalResult [[$A_SUCCESSOR_CONSTRAINT:__mlir_ods_local_successor_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$A_SUCCESSOR_CONSTRAINT:__mlir_ods_local_successor_constraint.*]]( // CHECK: if (!((successorPred(successor, *op)))) { // CHECK-NEXT: return op->emitOpError("successor #") << successorIndex << " ('" // CHECK-NEXT: << successorName << ")' failed to verify constraint: a successor"; @@ -94,13 +94,13 @@ def OpC : NS_Op<"op_c"> { // CHECK-NOT: << successorName << ")' failed to verify constraint: a successor"; /// Test that a successor constraint with a different description was generated. -// CHECK: static ::mlir::LogicalResult [[$O_SUCCESSOR_CONSTRAINT:__mlir_ods_local_successor_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$O_SUCCESSOR_CONSTRAINT:__mlir_ods_local_successor_constraint.*]]( // CHECK: if (!((successorPred(successor, *op)))) { // CHECK-NEXT: return op->emitOpError("successor #") << successorIndex << " ('" // CHECK-NEXT: << successorName << ")' failed to verify constraint: another successor"; /// Test that a region contraint was generated. -// CHECK: static ::mlir::LogicalResult [[$A_REGION_CONSTRAINT:__mlir_ods_local_region_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$A_REGION_CONSTRAINT:__mlir_ods_local_region_constraint.*]]( // CHECK: if (!((regionPred(region, *op)))) { // CHECK-NEXT: return op->emitOpError("region #") << regionIndex // CHECK-NEXT: << (regionName.empty() ? " " : " ('" + regionName + "') ") @@ -110,7 +110,7 @@ def OpC : NS_Op<"op_c"> { // CHECK-NOT: << "failed to verify constraint: a region"; /// Test that a region constraint with a different description was generated. -// CHECK: static ::mlir::LogicalResult [[$O_REGION_CONSTRAINT:__mlir_ods_local_region_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$O_REGION_CONSTRAINT:__mlir_ods_local_region_constraint.*]]( // CHECK: if (!((regionPred(region, *op)))) { // CHECK-NEXT: return op->emitOpError("region #") << regionIndex // CHECK-NEXT: << (regionName.empty() ? " " : " ('" + regionName + "') ") diff --git a/mlir/test/mlir-tblgen/interfaces-as-constraints.td b/mlir/test/mlir-tblgen/interfaces-as-constraints.td index 109bf07..b0438c2 100644 --- a/mlir/test/mlir-tblgen/interfaces-as-constraints.td +++ b/mlir/test/mlir-tblgen/interfaces-as-constraints.td @@ -23,22 +23,22 @@ def OpUsingAllOfThose : Op<Test_Dialect, "OpUsingAllOfThose"> { let results = (outs TopLevelTypeInterface:$res1, TypeInterfaceInNamespace:$res2); } -// CHECK: static ::mlir::LogicalResult {{__mlir_ods_local_type_constraint.*}}( +// CHECK: static ::llvm::LogicalResult {{__mlir_ods_local_type_constraint.*}}( // CHECK: if (!((::llvm::isa<TopLevelTypeInterface>(type)))) { // CHECK-NEXT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NEXT: << " must be TopLevelTypeInterface instance, but got " << type; -// CHECK: static ::mlir::LogicalResult {{__mlir_ods_local_type_constraint.*}}( +// CHECK: static ::llvm::LogicalResult {{__mlir_ods_local_type_constraint.*}}( // CHECK: if (!((::llvm::isa<test::TypeInterfaceInNamespace>(type)))) { // CHECK-NEXT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NEXT: << " must be TypeInterfaceInNamespace instance, but got " << type; -// CHECK: static ::mlir::LogicalResult {{__mlir_ods_local_attr_constraint.*}}( +// CHECK: static ::llvm::LogicalResult {{__mlir_ods_local_attr_constraint.*}}( // CHECK: if (attr && !((::llvm::isa<TopLevelAttrInterface>(attr)))) // CHECK-NEXT: return emitError() << "attribute '" << attrName // CHECK-NEXT: << "' failed to satisfy constraint: TopLevelAttrInterface instance"; -// CHECK: static ::mlir::LogicalResult {{__mlir_ods_local_attr_constraint.*}}( +// CHECK: static ::llvm::LogicalResult {{__mlir_ods_local_attr_constraint.*}}( // CHECK: if (attr && !((::llvm::isa<test::AttrInterfaceInNamespace>(attr)))) // CHECK-NEXT: return emitError() << "attribute '" << attrName // CHECK-NEXT: << "' failed to satisfy constraint: AttrInterfaceInNamespace instance"; diff --git a/mlir/test/mlir-tblgen/op-attribute.td b/mlir/test/mlir-tblgen/op-attribute.td index 6f2d430..a82cd2e 100644 --- a/mlir/test/mlir-tblgen/op-attribute.td +++ b/mlir/test/mlir-tblgen/op-attribute.td @@ -68,7 +68,7 @@ def AOp : NS_Op<"a_op", []> { // Test verify method // --- -// DEF: ::mlir::LogicalResult AOpAdaptor::verify +// DEF: ::llvm::LogicalResult AOpAdaptor::verify // DEF: ::mlir::Attribute tblgen_aAttr; // DEF: while (true) { // DEF-NEXT: if (namedAttrIt == namedAttrRange.end()) @@ -220,7 +220,7 @@ def AgetOp : Op<Test2_Dialect, "a_get_op", []> { // Test verify method // --- -// DEF: ::mlir::LogicalResult AgetOpAdaptor::verify +// DEF: ::llvm::LogicalResult AgetOpAdaptor::verify // DEF: ::mlir::Attribute tblgen_aAttr; // DEF: while (true) // DEF: ::mlir::Attribute tblgen_bAttr; diff --git a/mlir/test/mlir-tblgen/op-decl-and-defs.td b/mlir/test/mlir-tblgen/op-decl-and-defs.td index 836ddca..f6fe450 100644 --- a/mlir/test/mlir-tblgen/op-decl-and-defs.td +++ b/mlir/test/mlir-tblgen/op-decl-and-defs.td @@ -84,7 +84,7 @@ def NS_AOp : NS_Op<"a_op", [IsolatedFromAbove, IsolatedFromAbove]> { // CHECK: class AOpAdaptor : public AOpGenericAdaptor<::mlir::ValueRange> { // CHECK: public: // CHECK: AOpAdaptor(AOp -// CHECK: ::mlir::LogicalResult verify( +// CHECK: ::llvm::LogicalResult verify( // CHECK: }; // CHECK: class AOp : public ::mlir::Op<AOp, ::mlir::OpTrait::AtLeastNRegions<1>::Impl, ::mlir::OpTrait::AtLeastNResults<1>::Impl, ::mlir::OpTrait::ZeroSuccessors, ::mlir::OpTrait::AtLeastNOperands<1>::Impl, ::mlir::OpTrait::OpInvariants, ::mlir::OpTrait::IsIsolatedFromAbove @@ -120,9 +120,9 @@ def NS_AOp : NS_Op<"a_op", [IsolatedFromAbove, IsolatedFromAbove]> { // CHECK: static void build(::mlir::OpBuilder &, ::mlir::OperationState &odsState, ::mlir::TypeRange resultTypes, ::mlir::ValueRange operands, ::llvm::ArrayRef<::mlir::NamedAttribute> attributes, unsigned numRegions) // CHECK: static ::mlir::ParseResult parse(::mlir::OpAsmParser &parser, ::mlir::OperationState &result); // CHECK: void print(::mlir::OpAsmPrinter &p); -// CHECK: ::mlir::LogicalResult verifyInvariants(); +// CHECK: ::llvm::LogicalResult verifyInvariants(); // CHECK: static void getCanonicalizationPatterns(::mlir::RewritePatternSet &results, ::mlir::MLIRContext *context); -// CHECK: ::mlir::LogicalResult fold(FoldAdaptor adaptor, ::llvm::SmallVectorImpl<::mlir::OpFoldResult> &results); +// CHECK: ::llvm::LogicalResult fold(FoldAdaptor adaptor, ::llvm::SmallVectorImpl<::mlir::OpFoldResult> &results); // CHECK: // Display a graph for debugging purposes. // CHECK: void displayGraph(); // CHECK: }; @@ -205,7 +205,7 @@ def NS_FOp : NS_Op<"op_with_all_types_constraint", } // CHECK-LABEL: class FOp : -// CHECK: static ::mlir::LogicalResult inferReturnTypes +// CHECK: static ::llvm::LogicalResult inferReturnTypes def NS_GOp : NS_Op<"op_with_fixed_return_type", []> { let arguments = (ins AnyType:$a); @@ -213,7 +213,7 @@ def NS_GOp : NS_Op<"op_with_fixed_return_type", []> { } // CHECK-LABEL: class GOp : -// CHECK: static ::mlir::LogicalResult inferReturnTypes +// CHECK: static ::llvm::LogicalResult inferReturnTypes // Check default value for collective params builder. Check that other builders // are generated as well. diff --git a/mlir/test/mlir-tblgen/predicate.td b/mlir/test/mlir-tblgen/predicate.td index 44f395d..c1fcd3f 100644 --- a/mlir/test/mlir-tblgen/predicate.td +++ b/mlir/test/mlir-tblgen/predicate.td @@ -16,7 +16,7 @@ def OpA : NS_Op<"op_for_CPred_containing_multiple_same_placeholder", []> { let results = (outs Variadic<I32OrF32>:$y); } -// CHECK: static ::mlir::LogicalResult [[$INTEGER_FLOAT_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$INTEGER_FLOAT_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( // CHECK: if (!((type.isInteger(32) || type.isF32()))) { // CHECK-NEXT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NEXT: << " must be 32-bit integer or floating-point type, but got " << type; @@ -26,12 +26,12 @@ def OpA : NS_Op<"op_for_CPred_containing_multiple_same_placeholder", []> { // CHECK-NOT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NOT. << " must be 32-bit integer or floating-point type, but got " << type; -// CHECK: static ::mlir::LogicalResult [[$TENSOR_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$TENSOR_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( // CHECK: if (!(((::llvm::isa<::mlir::TensorType>(type))) && ([](::mlir::Type elementType) { return (true); }(::llvm::cast<::mlir::ShapedType>(type).getElementType())))) { // CHECK-NEXT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NEXT: << " must be tensor of any type values, but got " << type; -// CHECK: static ::mlir::LogicalResult [[$TENSOR_INTEGER_FLOAT_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$TENSOR_INTEGER_FLOAT_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( // CHECK: if (!(((::llvm::isa<::mlir::TensorType>(type))) && ([](::mlir::Type elementType) { return ((elementType.isF32())) || ((elementType.isSignlessInteger(32))); }(::llvm::cast<::mlir::ShapedType>(type).getElementType())))) { // CHECK-NEXT: return op->emitOpError(valueKind) << " #" << valueIndex // CHECK-NEXT: << " must be tensor of 32-bit float or 32-bit signless integer values, but got " << type; diff --git a/mlir/test/mlir-tblgen/rewriter-static-matcher.td b/mlir/test/mlir-tblgen/rewriter-static-matcher.td index 7a84dfd..c5debf5 100644 --- a/mlir/test/mlir-tblgen/rewriter-static-matcher.td +++ b/mlir/test/mlir-tblgen/rewriter-static-matcher.td @@ -48,11 +48,11 @@ def Foo : NativeCodeCall<"foo($_builder, $0)">; // Test static matcher for duplicate DagNode // --- -// CHECK: static ::mlir::LogicalResult [[$TYPE_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$TYPE_CONSTRAINT:__mlir_ods_local_type_constraint.*]]( // CHECK-NEXT: {{.*::mlir::Type type}} -// CHECK: static ::mlir::LogicalResult [[$ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( +// CHECK: static ::llvm::LogicalResult [[$ATTR_CONSTRAINT:__mlir_ods_local_attr_constraint.*]]( // CHECK-NEXT: {{.*::mlir::Attribute attr}} -// CHECK: static ::mlir::LogicalResult [[$DAG_MATCHER:static_dag_matcher.*]]( +// CHECK: static ::llvm::LogicalResult [[$DAG_MATCHER:static_dag_matcher.*]]( // CHECK: if(::mlir::failed([[$ATTR_CONSTRAINT]] // CHECK: if(::mlir::failed([[$TYPE_CONSTRAINT]] diff --git a/mlir/test/mlir-tblgen/typedefs.td b/mlir/test/mlir-tblgen/typedefs.td index 705da4e..b9e3a79 100644 --- a/mlir/test/mlir-tblgen/typedefs.td +++ b/mlir/test/mlir-tblgen/typedefs.td @@ -70,7 +70,7 @@ def B_CompoundTypeA : TestType<"CompoundA"> { // DECL-LABEL: class CompoundAType : public ::mlir::Type // DECL: static CompoundAType getChecked(::llvm::function_ref<::mlir::InFlightDiagnostic()> emitError, ::mlir::MLIRContext *context, int widthOfSomething, ::test::SimpleTypeA exampleTdType, SomeCppStruct exampleCppType, ::llvm::ArrayRef<int> dims, ::mlir::Type inner); -// DECL: static ::mlir::LogicalResult verify(::llvm::function_ref<::mlir::InFlightDiagnostic()> emitError, int widthOfSomething, ::test::SimpleTypeA exampleTdType, SomeCppStruct exampleCppType, ::llvm::ArrayRef<int> dims, ::mlir::Type inner); +// DECL: static ::llvm::LogicalResult verify(::llvm::function_ref<::mlir::InFlightDiagnostic()> emitError, int widthOfSomething, ::test::SimpleTypeA exampleTdType, SomeCppStruct exampleCppType, ::llvm::ArrayRef<int> dims, ::mlir::Type inner); // DECL: static constexpr ::llvm::StringLiteral getMnemonic() { // DECL: return {"cmpnd_a"}; // DECL: } diff --git a/mlir/test/python/python_test_ops.td b/mlir/test/python/python_test_ops.td index 9530198..c0bc184 100644 --- a/mlir/test/python/python_test_ops.td +++ b/mlir/test/python/python_test_ops.td @@ -128,7 +128,7 @@ def InferResultsOp : TestOp<"infer_results_op", [InferTypeOpInterface]> { let results = (outs AnyInteger:$single, AnyInteger:$doubled); let extraClassDeclaration = [{ - static ::mlir::LogicalResult inferReturnTypes( + static ::llvm::LogicalResult inferReturnTypes( ::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, ::mlir::OpaqueProperties, @@ -151,7 +151,7 @@ def InferResultsVariadicInputsOp : TestOp<"infer_results_variadic_inputs_op", let results = (outs I32OrF32:$res); let extraClassDeclaration = [{ - static ::mlir::LogicalResult inferReturnTypes( + static ::llvm::LogicalResult inferReturnTypes( ::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueRange operands, ::mlir::DictionaryAttr attributes, ::mlir::OpaqueProperties, @@ -180,7 +180,7 @@ def InferShapedTypeComponentsOp : TestOp<"infer_shaped_type_components_op", let results = (outs AnyTensor:$result); let extraClassDefinition = [{ - ::mlir::LogicalResult $cppClass::inferReturnTypeComponents( + ::llvm::LogicalResult $cppClass::inferReturnTypeComponents( ::mlir::MLIRContext *context, ::std::optional<::mlir::Location> location, ::mlir::ValueShapeRange operands, ::mlir::DictionaryAttr attributes, ::mlir::OpaqueProperties properties, ::mlir::RegionRange regions, diff --git a/mlir/tools/mlir-pdll-lsp-server/mlir-pdll-lsp-server.cpp b/mlir/tools/mlir-pdll-lsp-server/mlir-pdll-lsp-server.cpp index de50fd3..ffda7c8 100644 --- a/mlir/tools/mlir-pdll-lsp-server/mlir-pdll-lsp-server.cpp +++ b/mlir/tools/mlir-pdll-lsp-server/mlir-pdll-lsp-server.cpp @@ -6,7 +6,7 @@ // //===----------------------------------------------------------------------===// -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "mlir/Tools/mlir-pdll-lsp-server/MlirPdllLspServerMain.h" using namespace mlir; diff --git a/mlir/tools/mlir-src-sharder/mlir-src-sharder.cpp b/mlir/tools/mlir-src-sharder/mlir-src-sharder.cpp index b15f2ce..2f92ff2 100644 --- a/mlir/tools/mlir-src-sharder/mlir-src-sharder.cpp +++ b/mlir/tools/mlir-src-sharder/mlir-src-sharder.cpp @@ -7,7 +7,7 @@ //===----------------------------------------------------------------------===// #include "mlir/Support/FileUtilities.h" -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "llvm/Support/CommandLine.h" #include "llvm/Support/InitLLVM.h" #include "llvm/Support/MemoryBuffer.h" diff --git a/mlir/tools/mlir-tblgen/AttrOrTypeDefGen.cpp b/mlir/tools/mlir-tblgen/AttrOrTypeDefGen.cpp index 55bc071..ea0d152 100644 --- a/mlir/tools/mlir-tblgen/AttrOrTypeDefGen.cpp +++ b/mlir/tools/mlir-tblgen/AttrOrTypeDefGen.cpp @@ -297,7 +297,7 @@ void DefGen::emitBuilders() { void DefGen::emitVerifier() { defCls.declare<UsingDeclaration>("Base::getChecked"); defCls.declareStaticMethod( - "::mlir::LogicalResult", "verify", + "::llvm::LogicalResult", "verify", getBuilderParams({{"::llvm::function_ref<::mlir::InFlightDiagnostic()>", "emitError"}})); } @@ -819,7 +819,7 @@ void DefGenerator::emitParsePrintDispatch(ArrayRef<AttrOrTypeDef> defs) { strfmt("generated{0}Parser", valueType), Method::StaticInline, std::move(params)); // Declare the printer. - Method printer("::mlir::LogicalResult", + Method printer("::llvm::LogicalResult", strfmt("generated{0}Printer", valueType), Method::StaticInline, {{strfmt("::mlir::{0}", valueType), "def"}, {"::mlir::AsmPrinter &", "printer"}}); @@ -839,7 +839,7 @@ void DefGenerator::emitParsePrintDispatch(ArrayRef<AttrOrTypeDef> defs) { // The printer dispatch uses llvm::TypeSwitch to find and call the correct // printer. printer.body() << " return ::llvm::TypeSwitch<::mlir::" << valueType - << ", ::mlir::LogicalResult>(def)"; + << ", ::llvm::LogicalResult>(def)"; const char *const printValue = R"( .Case<{0}>([&](auto t) {{ printer << {0}::getMnemonic();{1} return ::mlir::success(); diff --git a/mlir/tools/mlir-tblgen/AttrOrTypeFormatGen.cpp b/mlir/tools/mlir-tblgen/AttrOrTypeFormatGen.cpp index 50378fe..ffd5a39 100644 --- a/mlir/tools/mlir-tblgen/AttrOrTypeFormatGen.cpp +++ b/mlir/tools/mlir-tblgen/AttrOrTypeFormatGen.cpp @@ -9,7 +9,6 @@ #include "AttrOrTypeFormatGen.h" #include "FormatGen.h" #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "mlir/TableGen/AttrOrTypeDef.h" #include "mlir/TableGen/Format.h" #include "mlir/TableGen/GenInfo.h" diff --git a/mlir/tools/mlir-tblgen/DialectGen.cpp b/mlir/tools/mlir-tblgen/DialectGen.cpp index 46e585a..2412876 100644 --- a/mlir/tools/mlir-tblgen/DialectGen.cpp +++ b/mlir/tools/mlir-tblgen/DialectGen.cpp @@ -166,7 +166,7 @@ static const char *const constantMaterializerDecl = R"( static const char *const opAttrVerifierDecl = R"( /// Provides a hook for verifying dialect attributes attached to the given /// op. - ::mlir::LogicalResult verifyOperationAttribute( + ::llvm::LogicalResult verifyOperationAttribute( ::mlir::Operation *op, ::mlir::NamedAttribute attribute) override; )"; @@ -174,7 +174,7 @@ static const char *const opAttrVerifierDecl = R"( static const char *const regionArgAttrVerifierDecl = R"( /// Provides a hook for verifying dialect attributes attached to the given /// op's region argument. - ::mlir::LogicalResult verifyRegionArgAttribute( + ::llvm::LogicalResult verifyRegionArgAttribute( ::mlir::Operation *op, unsigned regionIndex, unsigned argIndex, ::mlir::NamedAttribute attribute) override; )"; @@ -183,7 +183,7 @@ static const char *const regionArgAttrVerifierDecl = R"( static const char *const regionResultAttrVerifierDecl = R"( /// Provides a hook for verifying dialect attributes attached to the given /// op's region result. - ::mlir::LogicalResult verifyRegionResultAttribute( + ::llvm::LogicalResult verifyRegionResultAttribute( ::mlir::Operation *op, unsigned regionIndex, unsigned resultIndex, ::mlir::NamedAttribute attribute) override; )"; diff --git a/mlir/tools/mlir-tblgen/FormatGen.h b/mlir/tools/mlir-tblgen/FormatGen.h index b061d4d..1dc2cb3 100644 --- a/mlir/tools/mlir-tblgen/FormatGen.h +++ b/mlir/tools/mlir-tblgen/FormatGen.h @@ -15,7 +15,6 @@ #define MLIR_TOOLS_MLIRTBLGEN_FORMATGEN_H_ #include "mlir/Support/LLVM.h" -#include "mlir/Support/LogicalResult.h" #include "llvm/ADT/StringRef.h" #include "llvm/ADT/StringSet.h" #include "llvm/Support/Allocator.h" diff --git a/mlir/tools/mlir-tblgen/LLVMIRConversionGen.cpp b/mlir/tools/mlir-tblgen/LLVMIRConversionGen.cpp index 2c7acec..ebadfe4 100644 --- a/mlir/tools/mlir-tblgen/LLVMIRConversionGen.cpp +++ b/mlir/tools/mlir-tblgen/LLVMIRConversionGen.cpp @@ -11,7 +11,6 @@ // //===----------------------------------------------------------------------===// -#include "mlir/Support/LogicalResult.h" #include "mlir/TableGen/Argument.h" #include "mlir/TableGen/Attribute.h" #include "mlir/TableGen/GenInfo.h" diff --git a/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp b/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp index 3146e65..0fc750c 100644 --- a/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp +++ b/mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp @@ -1250,7 +1250,7 @@ void OpEmitter::genPropertiesSupport() { auto &setPropMethod = opClass .addStaticMethod( - "::mlir::LogicalResult", "setPropertiesFromAttr", + "::llvm::LogicalResult", "setPropertiesFromAttr", MethodParameter("Properties &", "prop"), MethodParameter("::mlir::Attribute", "attr"), MethodParameter( @@ -1292,7 +1292,7 @@ void OpEmitter::genPropertiesSupport() { auto &verifyInherentAttrsMethod = opClass .addStaticMethod( - "::mlir::LogicalResult", "verifyInherentAttrs", + "::llvm::LogicalResult", "verifyInherentAttrs", MethodParameter("::mlir::OperationName", "opName"), MethodParameter("::mlir::NamedAttrList &", "attrs"), MethodParameter( @@ -1597,7 +1597,7 @@ void OpEmitter::genPropertiesSupportForBytecode( ArrayRef<ConstArgument> attrOrProperties) { if (op.useCustomPropertiesEncoding()) { opClass.declareStaticMethod( - "::mlir::LogicalResult", "readProperties", + "::llvm::LogicalResult", "readProperties", MethodParameter("::mlir::DialectBytecodeReader &", "reader"), MethodParameter("::mlir::OperationState &", "state")); opClass.declareMethod( @@ -1609,7 +1609,7 @@ void OpEmitter::genPropertiesSupportForBytecode( auto &readPropertiesMethod = opClass .addStaticMethod( - "::mlir::LogicalResult", "readProperties", + "::llvm::LogicalResult", "readProperties", MethodParameter("::mlir::DialectBytecodeReader &", "reader"), MethodParameter("::mlir::OperationState &", "state")) ->body(); @@ -3226,7 +3226,7 @@ void OpEmitter::genCanonicalizerDecls() { SmallVector<MethodParameter> paramList; paramList.emplace_back(op.getCppClassName(), "op"); paramList.emplace_back("::mlir::PatternRewriter &", "rewriter"); - auto *m = opClass.declareStaticMethod("::mlir::LogicalResult", + auto *m = opClass.declareStaticMethod("::llvm::LogicalResult", "canonicalize", std::move(paramList)); ERROR_IF_PRUNED(m, "canonicalize", op); } @@ -3272,7 +3272,7 @@ void OpEmitter::genFolderDecls() { } else { paramList.emplace_back("::llvm::SmallVectorImpl<::mlir::OpFoldResult> &", "results"); - retType = "::mlir::LogicalResult"; + retType = "::llvm::LogicalResult"; } auto *m = opClass.declareMethod(retType, "fold", std::move(paramList)); @@ -3563,7 +3563,7 @@ void OpEmitter::genPrinter() { void OpEmitter::genVerifier() { auto *implMethod = - opClass.addMethod("::mlir::LogicalResult", "verifyInvariantsImpl"); + opClass.addMethod("::llvm::LogicalResult", "verifyInvariantsImpl"); ERROR_IF_PRUNED(implMethod, "verifyInvariantsImpl", op); auto &implBody = implMethod->body(); bool useProperties = emitHelper.hasProperties(); @@ -3592,7 +3592,7 @@ void OpEmitter::genVerifier() { // This may not act as their expectation because this doesn't call any // verifiers of native/interface traits. Needs to review those use cases and // see if we should use the mlir::verify() instead. - auto *method = opClass.addMethod("::mlir::LogicalResult", "verifyInvariants"); + auto *method = opClass.addMethod("::llvm::LogicalResult", "verifyInvariants"); ERROR_IF_PRUNED(method, "verifyInvariants", op); auto &body = method->body(); if (def.getValueAsBit("hasVerifier")) { @@ -3607,13 +3607,13 @@ void OpEmitter::genVerifier() { void OpEmitter::genCustomVerifier() { if (def.getValueAsBit("hasVerifier")) { - auto *method = opClass.declareMethod("::mlir::LogicalResult", "verify"); + auto *method = opClass.declareMethod("::llvm::LogicalResult", "verify"); ERROR_IF_PRUNED(method, "verify", op); } if (def.getValueAsBit("hasRegionVerifier")) { auto *method = - opClass.declareMethod("::mlir::LogicalResult", "verifyRegions"); + opClass.declareMethod("::llvm::LogicalResult", "verifyRegions"); ERROR_IF_PRUNED(method, "verifyRegions", op); } } @@ -4334,7 +4334,7 @@ OpOperandAdaptorEmitter::OpOperandAdaptorEmitter( } void OpOperandAdaptorEmitter::addVerification() { - auto *method = adaptor.addMethod("::mlir::LogicalResult", "verify", + auto *method = adaptor.addMethod("::llvm::LogicalResult", "verify", MethodParameter("::mlir::Location", "loc")); ERROR_IF_PRUNED(method, "verify", op); auto &body = method->body(); diff --git a/mlir/tools/mlir-tblgen/OpFormatGen.cpp b/mlir/tools/mlir-tblgen/OpFormatGen.cpp index f7cc0a2..a97d876 100644 --- a/mlir/tools/mlir-tblgen/OpFormatGen.cpp +++ b/mlir/tools/mlir-tblgen/OpFormatGen.cpp @@ -1200,7 +1200,7 @@ static void genParsedAttrPropertiesSetter(OperationFormat &fmt, Operator &op, paramList.emplace_back("::llvm::function_ref<::mlir::InFlightDiagnostic()>", "emitError"); - Method *method = opClass.addStaticMethod("::mlir::LogicalResult", + Method *method = opClass.addStaticMethod("::llvm::LogicalResult", "setPropertiesFromParsedAttr", std::move(paramList)); MethodBody &body = method->body().indent(); diff --git a/mlir/tools/mlir-tblgen/OpInterfacesGen.cpp b/mlir/tools/mlir-tblgen/OpInterfacesGen.cpp index 2a7406f..4b06b92 100644 --- a/mlir/tools/mlir-tblgen/OpInterfacesGen.cpp +++ b/mlir/tools/mlir-tblgen/OpInterfacesGen.cpp @@ -480,7 +480,7 @@ void InterfaceGenerator::emitTraitDecl(const Interface &interface, tblgen::FmtContext verifyCtx; verifyCtx.addSubst("_op", "op"); os << llvm::formatv( - " static ::mlir::LogicalResult {0}(::mlir::Operation *op) ", + " static ::llvm::LogicalResult {0}(::mlir::Operation *op) ", (interface.verifyWithRegions() ? "verifyRegionTrait" : "verifyTrait")) << "{\n " << tblgen::tgfmt(verify->trim(), &verifyCtx) diff --git a/mlir/tools/mlir-tblgen/RewriterGen.cpp b/mlir/tools/mlir-tblgen/RewriterGen.cpp index 0d3b29c..2c79ba2 100644 --- a/mlir/tools/mlir-tblgen/RewriterGen.cpp +++ b/mlir/tools/mlir-tblgen/RewriterGen.cpp @@ -366,7 +366,7 @@ std::string PatternEmitter::handleConstantAttr(Attribute attr, void PatternEmitter::emitStaticMatcher(DagNode tree, std::string funcName) { os << formatv( - "static ::mlir::LogicalResult {0}(::mlir::PatternRewriter &rewriter, " + "static ::llvm::LogicalResult {0}(::mlir::PatternRewriter &rewriter, " "::mlir::Operation *op0, ::llvm::SmallVector<::mlir::Operation " "*, 4> &tblgen_ops", funcName); @@ -1081,7 +1081,7 @@ void PatternEmitter::emit(StringRef rewriteName) { { auto classScope = os.scope(); os.printReindented(R"( - ::mlir::LogicalResult matchAndRewrite(::mlir::Operation *op0, + ::llvm::LogicalResult matchAndRewrite(::mlir::Operation *op0, ::mlir::PatternRewriter &rewriter) const override {)") << '\n'; { diff --git a/mlir/tools/mlir-translate/mlir-translate.cpp b/mlir/tools/mlir-translate/mlir-translate.cpp index 309def8..9d2dd88 100644 --- a/mlir/tools/mlir-translate/mlir-translate.cpp +++ b/mlir/tools/mlir-translate/mlir-translate.cpp @@ -12,7 +12,7 @@ //===----------------------------------------------------------------------===// #include "mlir/InitAllTranslations.h" -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "mlir/Tools/mlir-translate/MlirTranslateMain.h" using namespace mlir; diff --git a/mlir/tools/mlir-vulkan-runner/VulkanRuntime.h b/mlir/tools/mlir-vulkan-runner/VulkanRuntime.h index 9fa52b0..17db8d0 100644 --- a/mlir/tools/mlir-vulkan-runner/VulkanRuntime.h +++ b/mlir/tools/mlir-vulkan-runner/VulkanRuntime.h @@ -13,7 +13,7 @@ #ifndef VULKAN_RUNTIME_H #define VULKAN_RUNTIME_H -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include <unordered_map> #include <vector> diff --git a/mlir/tools/tblgen-lsp-server/tblgen-lsp-server.cpp b/mlir/tools/tblgen-lsp-server/tblgen-lsp-server.cpp index 90b2d98..e8fdbfc 100644 --- a/mlir/tools/tblgen-lsp-server/tblgen-lsp-server.cpp +++ b/mlir/tools/tblgen-lsp-server/tblgen-lsp-server.cpp @@ -6,7 +6,7 @@ // //===----------------------------------------------------------------------===// -#include "mlir/Support/LogicalResult.h" +#include "mlir/Support/LLVM.h" #include "mlir/Tools/tblgen-lsp-server/TableGenLspServerMain.h" using namespace mlir; diff --git a/mlir/unittests/Analysis/Presburger/ParserTest.cpp b/mlir/unittests/Analysis/Presburger/ParserTest.cpp index 4c9f54f..06b728c 100644 --- a/mlir/unittests/Analysis/Presburger/ParserTest.cpp +++ b/mlir/unittests/Analysis/Presburger/ParserTest.cpp @@ -45,6 +45,18 @@ static bool parseAndCompare(StringRef str, const IntegerPolyhedron &ex) { } TEST(ParseFACTest, ParseAndCompareTest) { + // constant-fold addition + EXPECT_TRUE(parseAndCompare("() : (4 + 3 >= 0)", + makeFACFromConstraints(0, 0, {}, {}))); + + // constant-fold addition + multiplication + EXPECT_TRUE(parseAndCompare("()[a] : (4 * 3 == 10 + 2)", + makeFACFromConstraints(0, 1, {}, {}))); + + // constant-fold ceildiv + floordiv + EXPECT_TRUE(parseAndCompare("(x) : (11 ceildiv 3 == 13 floordiv 3)", + makeFACFromConstraints(1, 0, {}, {}))); + // simple ineq EXPECT_TRUE(parseAndCompare("(x)[] : (x >= 0)", makeFACFromConstraints(1, 0, {{1, 0}}))); @@ -57,6 +69,11 @@ TEST(ParseFACTest, ParseAndCompareTest) { EXPECT_TRUE(parseAndCompare("(x)[] : (7 * x >= 0, -7 * x + 5 >= 0)", makeFACFromConstraints(1, 0, {{7, 0}, {-7, 5}}))); + // multiplication distribution + EXPECT_TRUE( + parseAndCompare("(x) : (2 * x >= 2, (-7 + x * 9) * 5 >= 0)", + makeFACFromConstraints(1, 0, {{2, -2}, {45, -35}}))); + // multiple dimensions EXPECT_TRUE(parseAndCompare("(x,y,z)[] : (x + y - z >= 0)", makeFACFromConstraints(3, 0, {{1, 1, -1, 0}}))); @@ -70,20 +87,61 @@ TEST(ParseFACTest, ParseAndCompareTest) { EXPECT_TRUE(parseAndCompare("()[a] : (2 * a - 4 == 0)", makeFACFromConstraints(0, 1, {}, {{2, -4}}))); + // no linear terms + EXPECT_TRUE(parseAndCompare( + "(x, y) : (26 * (x floordiv 6) == y floordiv 3)", + makeFACFromConstraints(2, 0, {}, {{0, 0, 26, -1, 0}}, + {{{1, 0, 0}, 6}, {{0, 1, 0, 0}, 3}}))); + // simple floordiv EXPECT_TRUE(parseAndCompare( "(x, y) : (y - 3 * ((x + y - 13) floordiv 3) - 42 == 0)", makeFACFromConstraints(2, 0, {}, {{0, 1, -3, -42}}, {{{1, 1, -13}, 3}}))); + // simple ceildiv + EXPECT_TRUE(parseAndCompare( + "(x, y) : (y - 3 * ((x + y - 13) ceildiv 3) - 42 == 0)", + makeFACFromConstraints(2, 0, {}, {{0, 1, -3, -42}}, {{{1, 1, -11}, 3}}))); + + // simple mod + EXPECT_TRUE(parseAndCompare( + "(x, y) : (y - 3 * ((x + y - 13) mod 3) - 42 == 0)", + makeFACFromConstraints(2, 0, {}, {{-3, -2, 9, -3}}, {{{1, 1, -13}, 3}}))); + // multiple floordiv EXPECT_TRUE(parseAndCompare( "(x, y) : (y - x floordiv 3 - y floordiv 2 == 0)", makeFACFromConstraints(2, 0, {}, {{0, 1, -1, -1, 0}}, {{{1, 0, 0}, 3}, {{0, 1, 0, 0}, 2}}))); + // multiple ceildiv + EXPECT_TRUE(parseAndCompare( + "(x, y) : (y - x ceildiv 3 - y ceildiv 2 == 0)", + makeFACFromConstraints(2, 0, {}, {{0, 1, -1, -1, 0}}, + {{{1, 0, 2}, 3}, {{0, 1, 0, 1}, 2}}))); + + // multiple mod + EXPECT_TRUE(parseAndCompare( + "(x, y) : (y - x mod 3 - y mod 2 == 0)", + makeFACFromConstraints(2, 0, {}, {{-1, 0, 3, 2, 0}}, + {{{1, 0, 0}, 3}, {{0, 1, 0, 0}, 2}}))); + // nested floordiv EXPECT_TRUE(parseAndCompare( "(x, y) : (y - (x + y floordiv 2) floordiv 3 == 0)", makeFACFromConstraints(2, 0, {}, {{0, 1, 0, -1, 0}}, {{{0, 1, 0}, 2}, {{1, 0, 1, 0}, 3}}))); + + // nested mod + EXPECT_TRUE(parseAndCompare( + "(x, y) : (y - (x + y mod 2) mod 3 == 0)", + makeFACFromConstraints(2, 0, {}, {{-1, 0, 2, 3, 0}}, + {{{0, 1, 0}, 2}, {{1, 1, -2, 0}, 3}}))); + + // nested floordiv + ceildiv + mod + EXPECT_TRUE(parseAndCompare( + "(x, y) : ((2 * x + 3 * (y floordiv 2) + x mod 7 + 1) ceildiv 3 == 42)", + makeFACFromConstraints( + 2, 0, {}, {{0, 0, 0, 0, 1, -42}}, + {{{0, 1, 0}, 2}, {{1, 0, 0, 0}, 7}, {{3, 0, 3, -7, 3}, 3}}))); } diff --git a/mlir/unittests/Rewrite/PatternBenefit.cpp b/mlir/unittests/Rewrite/PatternBenefit.cpp index 79552f0..65ea4ee 100644 --- a/mlir/unittests/Rewrite/PatternBenefit.cpp +++ b/mlir/unittests/Rewrite/PatternBenefit.cpp @@ -27,7 +27,7 @@ TEST(PatternBenefitTest, BenefitOrder) { Pattern1(mlir::MLIRContext *context, bool *called) : OpRewritePattern<ModuleOp>(context, /*benefit*/ 1), called(called) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(ModuleOp /*op*/, mlir::PatternRewriter & /*rewriter*/) const override { *called = true; @@ -43,7 +43,7 @@ TEST(PatternBenefitTest, BenefitOrder) { : RewritePattern(MatchAnyOpTypeTag(), /*benefit=*/2, context), called(called) {} - mlir::LogicalResult + llvm::LogicalResult matchAndRewrite(Operation * /*op*/, mlir::PatternRewriter & /*rewriter*/) const override { *called = true; diff --git a/openmp/runtime/src/dllexports b/openmp/runtime/src/dllexports index 747b828..0667d53 100644 --- a/openmp/runtime/src/dllexports +++ b/openmp/runtime/src/dllexports @@ -1276,5 +1276,6 @@ kmp_set_disp_num_buffers 890 %endif __kmpc_set_thread_limit +__kmpc_dispatch_deinit # end of file # diff --git a/utils/bazel/llvm-project-overlay/clang/BUILD.bazel b/utils/bazel/llvm-project-overlay/clang/BUILD.bazel index 725ac6b..caebcd0 100644 --- a/utils/bazel/llvm-project-overlay/clang/BUILD.bazel +++ b/utils/bazel/llvm-project-overlay/clang/BUILD.bazel @@ -1869,11 +1869,19 @@ cc_library( cc_library( name = "interpreter", - srcs = glob([ - "lib/Interpreter/*.cpp", - "lib/Interpreter/*.h", - ]), - hdrs = glob(["include/clang/Interpreter/*.h"]), + srcs = glob( + [ + "lib/Interpreter/*.cpp", + "lib/Interpreter/*.h", + ], + exclude = ["lib/Interpreter/Wasm.cpp"], + ), + hdrs = glob( + [ + "include/clang/Interpreter/*.h", + ], + exclude = ["lib/Interpreter/Wasm.cpp"], + ), includes = ["include"], deps = [ ":analysis", |